diff --git a/.gitattributes b/.gitattributes index 30954f4fb8537ec6a956d0a2447e78b3982b441a..b5fce5c0404eba8ddeee667b4a115d96cd902956 100644 --- a/.gitattributes +++ b/.gitattributes @@ -38,3 +38,7 @@ comfyui_controlnet_aux/examples/ExecuteAll.png filter=lfs diff=lfs merge=lfs -te comfyui_controlnet_aux/examples/ExecuteAll1.jpg filter=lfs diff=lfs merge=lfs -text comfyui_controlnet_aux/examples/ExecuteAll2.jpg filter=lfs diff=lfs merge=lfs -text comfyui_controlnet_aux/src/custom_controlnet_aux/mesh_graphormer/hand_landmarker.task filter=lfs diff=lfs merge=lfs -text +ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d0.mp4 filter=lfs diff=lfs merge=lfs -text +ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d3.mp4 filter=lfs diff=lfs merge=lfs -text +ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d6.mp4 filter=lfs diff=lfs merge=lfs -text +ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d9.mp4 filter=lfs diff=lfs merge=lfs -text diff --git a/ComfyUI-AdvancedLivePortrait/.github/workflows/publish.yml b/ComfyUI-AdvancedLivePortrait/.github/workflows/publish.yml new file mode 100644 index 0000000000000000000000000000000000000000..90c00b5c5c2a1da4918695ee93b0fa3b72e7836c --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/.github/workflows/publish.yml @@ -0,0 +1,24 @@ +name: Publish to Comfy registry +on: + workflow_dispatch: + push: + branches: + - main + - master + paths: + - "pyproject.toml" + +jobs: + publish-node: + name: Publish Custom Node to registry + runs-on: ubuntu-latest + # if this is a forked repository. Skipping the workflow. + if: github.event.repository.fork == false + steps: + - name: Check out code + uses: actions/checkout@v4 + - name: Publish Custom Node + uses: Comfy-Org/publish-node-action@main + with: + ## Add your own personal access token to your Github Repository secrets and reference it here. + personal_access_token: ${{ secrets.REGISTRY_ACCESS_TOKEN }} diff --git a/ComfyUI-AdvancedLivePortrait/.gitignore b/ComfyUI-AdvancedLivePortrait/.gitignore new file mode 100644 index 0000000000000000000000000000000000000000..4dbbb2abcd52fddff617738408552d5985601b9d --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/.gitignore @@ -0,0 +1,2 @@ +__pycache__ +.idea \ No newline at end of file diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/__pycache__/live_portrait_wrapper.cpython-312.pyc b/ComfyUI-AdvancedLivePortrait/LivePortrait/__pycache__/live_portrait_wrapper.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..638792540bea122b249c770a395bf90cc9f4285a Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/LivePortrait/__pycache__/live_portrait_wrapper.cpython-312.pyc differ diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/config/__pycache__/inference_config.cpython-312.pyc b/ComfyUI-AdvancedLivePortrait/LivePortrait/config/__pycache__/inference_config.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..eb2963abf061d31a390fb701869755fc205c14cc Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/LivePortrait/config/__pycache__/inference_config.cpython-312.pyc differ diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/config/inference_config.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/config/inference_config.py new file mode 100644 index 0000000000000000000000000000000000000000..b08316d19bd1a9cec975252a8c4bae3de8b6c62e --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/LivePortrait/config/inference_config.py @@ -0,0 +1,8 @@ + +import os + +current_file_path = os.path.abspath(__file__) +current_directory = os.path.dirname(current_file_path) +class InferenceConfig: + def __init__(self): + self.flag_use_half_precision: bool = False # whether to use half precision \ No newline at end of file diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/config/models.yaml b/ComfyUI-AdvancedLivePortrait/LivePortrait/config/models.yaml new file mode 100644 index 0000000000000000000000000000000000000000..75efb6aa8d8f676611656837f74b74a1583c6d5c --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/LivePortrait/config/models.yaml @@ -0,0 +1,43 @@ +model_params: + appearance_feature_extractor_params: # the F in the paper + image_channel: 3 + block_expansion: 64 + num_down_blocks: 2 + max_features: 512 + reshape_channel: 32 + reshape_depth: 16 + num_resblocks: 6 + motion_extractor_params: # the M in the paper + num_kp: 21 + backbone: convnextv2_tiny + warping_module_params: # the W in the paper + num_kp: 21 + block_expansion: 64 + max_features: 512 + num_down_blocks: 2 + reshape_channel: 32 + estimate_occlusion_map: True + dense_motion_params: + block_expansion: 32 + max_features: 1024 + num_blocks: 5 + reshape_depth: 16 + compress: 4 + spade_generator_params: # the G in the paper + upscale: 2 # represents upsample factor 256x256 -> 512x512 + block_expansion: 64 + max_features: 512 + num_down_blocks: 2 + stitching_retargeting_module_params: # the S in the paper + stitching: + input_size: 126 # (21*3)*2 + hidden_sizes: [128, 128, 64] + output_size: 65 # (21*3)+2(tx,ty) + lip: + input_size: 65 # (21*3)+2 + hidden_sizes: [128, 128, 64] + output_size: 63 # (21*3) + eye: + input_size: 66 # (21*3)+3 + hidden_sizes: [256, 256, 128, 128, 64] + output_size: 63 # (21*3) diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/live_portrait_wrapper.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/live_portrait_wrapper.py new file mode 100644 index 0000000000000000000000000000000000000000..b657917bcd7cc740fe63555fd56b6f1960315871 --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/LivePortrait/live_portrait_wrapper.py @@ -0,0 +1,150 @@ +import numpy as np +import torch + +from .utils.helper import concat_feat +from .utils.camera import headpose_pred_to_degree, get_rotation_matrix +from .config.inference_config import InferenceConfig + +class LivePortraitWrapper(object): + + def __init__(self, cfg: InferenceConfig, appearance_feature_extractor, motion_extractor, + warping_module, spade_generator, stitching_retargeting_module): + + self.appearance_feature_extractor = appearance_feature_extractor + self.motion_extractor = motion_extractor + self.warping_module = warping_module + self.spade_generator = spade_generator + self.stitching_retargeting_module = stitching_retargeting_module + + self.cfg = cfg + + def extract_feature_3d(self, x: torch.Tensor) -> torch.Tensor: + """ get the appearance feature of the image by F + x: Bx3xHxW, normalized to 0~1 + """ + with torch.no_grad(): + feature_3d = self.appearance_feature_extractor(x) + + return feature_3d.float() + + def get_kp_info(self, x: torch.Tensor, **kwargs) -> dict: + """ get the implicit keypoint information + x: Bx3xHxW, normalized to 0~1 + flag_refine_info: whether to trandform the pose to degrees and the dimention of the reshape + return: A dict contains keys: 'pitch', 'yaw', 'roll', 't', 'exp', 'scale', 'kp' + """ + with torch.no_grad(): + kp_info = self.motion_extractor(x) + + if self.cfg.flag_use_half_precision: + # float the dict + for k, v in kp_info.items(): + if isinstance(v, torch.Tensor): + kp_info[k] = v.float() + + flag_refine_info: bool = kwargs.get('flag_refine_info', True) + if flag_refine_info: + bs = kp_info['kp'].shape[0] + kp_info['pitch'] = headpose_pred_to_degree(kp_info['pitch'])[:, None] # Bx1 + kp_info['yaw'] = headpose_pred_to_degree(kp_info['yaw'])[:, None] # Bx1 + kp_info['roll'] = headpose_pred_to_degree(kp_info['roll'])[:, None] # Bx1 + kp_info['kp'] = kp_info['kp'].reshape(bs, -1, 3) # BxNx3 + kp_info['exp'] = kp_info['exp'].reshape(bs, -1, 3) # BxNx3 + + return kp_info + def transform_keypoint(self, kp_info: dict): + """ + transform the implicit keypoints with the pose, shift, and expression deformation + kp: BxNx3 + """ + kp = kp_info['kp'] # (bs, k, 3) + pitch, yaw, roll = kp_info['pitch'], kp_info['yaw'], kp_info['roll'] + + t, exp = kp_info['t'], kp_info['exp'] + scale = kp_info['scale'] + + pitch = headpose_pred_to_degree(pitch) + yaw = headpose_pred_to_degree(yaw) + roll = headpose_pred_to_degree(roll) + + bs = kp.shape[0] + if kp.ndim == 2: + num_kp = kp.shape[1] // 3 # Bx(num_kpx3) + else: + num_kp = kp.shape[1] # Bxnum_kpx3 + + rot_mat = get_rotation_matrix(pitch, yaw, roll) # (bs, 3, 3) + + # Eqn.2: s * (R * x_c,s + exp) + t + kp_transformed = kp.view(bs, num_kp, 3) @ rot_mat + exp.view(bs, num_kp, 3) + kp_transformed *= scale[..., None] # (bs, k, 3) * (bs, 1, 1) = (bs, k, 3) + kp_transformed[:, :, 0:2] += t[:, None, 0:2] # remove z, only apply tx ty + + return kp_transformed + + def stitch(self, kp_source: torch.Tensor, kp_driving: torch.Tensor) -> torch.Tensor: + """ + kp_source: BxNx3 + kp_driving: BxNx3 + Return: Bx(3*num_kp+2) + """ + feat_stiching = concat_feat(kp_source, kp_driving) + + with torch.no_grad(): + delta = self.stitching_retargeting_module['stitching'](feat_stiching) + + return delta + + def stitching(self, kp_source: torch.Tensor, kp_driving: torch.Tensor) -> torch.Tensor: + """ conduct the stitching + kp_source: Bxnum_kpx3 + kp_driving: Bxnum_kpx3 + """ + + if self.stitching_retargeting_module is not None: + + bs, num_kp = kp_source.shape[:2] + + kp_driving_new = kp_driving.clone() + delta = self.stitch(kp_source, kp_driving_new) + + delta_exp = delta[..., :3*num_kp].reshape(bs, num_kp, 3) # 1x20x3 + delta_tx_ty = delta[..., 3*num_kp:3*num_kp+2].reshape(bs, 1, 2) # 1x1x2 + + kp_driving_new += delta_exp + kp_driving_new[..., :2] += delta_tx_ty + + return kp_driving_new + + return kp_driving + + def warp_decode(self, feature_3d: torch.Tensor, kp_source: torch.Tensor, kp_driving: torch.Tensor) -> torch.Tensor: + """ get the image after the warping of the implicit keypoints + feature_3d: Bx32x16x64x64, feature volume + kp_source: BxNx3 + kp_driving: BxNx3 + """ + # The line 18 in Algorithm 1: D(W(f_s; x_s, x′_d,i)） + with torch.no_grad(): + # get decoder input + ret_dct = self.warping_module(feature_3d, kp_source=kp_source, kp_driving=kp_driving) + # decode + ret_dct['out'] = self.spade_generator(feature=ret_dct['out']) + + # float the dict + if self.cfg.flag_use_half_precision: + for k, v in ret_dct.items(): + if isinstance(v, torch.Tensor): + ret_dct[k] = v.float() + + return ret_dct + + def parse_output(self, out: torch.Tensor) -> np.ndarray: + """ construct the output as standard + return: 1xHxWx3, uint8 + """ + out = np.transpose(out.data.cpu().numpy(), [0, 2, 3, 1]) # 1x3xHxW -> 1xHxWx3 + out = np.clip(out, 0, 1) # clip to 0~1 + out = np.clip(out * 255, 0, 255).astype(np.uint8) # 0~1 -> 0~255 + + return out diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__init__.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/__init__.cpython-312.pyc b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/__init__.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..869cdd2daf0db2b6f4838573292d9b246db6e891 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/__init__.cpython-312.pyc differ diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/appearance_feature_extractor.cpython-312.pyc b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/appearance_feature_extractor.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..389def4ba20359970ff54026fb0f1d1c09848cd3 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/appearance_feature_extractor.cpython-312.pyc differ diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/convnextv2.cpython-312.pyc b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/convnextv2.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..087533f97b81cfa68da658e0d30715cb22105067 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/convnextv2.cpython-312.pyc differ diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/dense_motion.cpython-312.pyc b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/dense_motion.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..a12216cc13d0b51b371d448cdd48edab55a1ae23 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/dense_motion.cpython-312.pyc differ diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/motion_extractor.cpython-312.pyc b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/motion_extractor.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..291ff3a516422dcc82cf3055a2f002a8b1744058 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/motion_extractor.cpython-312.pyc differ diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/spade_generator.cpython-312.pyc b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/spade_generator.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..4f3ecfa363419fce2fed500dbafb1dd4461fbb66 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/spade_generator.cpython-312.pyc differ diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/stitching_retargeting_network.cpython-312.pyc b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/stitching_retargeting_network.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..7e71536791ac833ad2aa02377f1b2c0ffb5789db Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/stitching_retargeting_network.cpython-312.pyc differ diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/util.cpython-312.pyc b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/util.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..7a0cdae2d93c79a2c618fd0469f9234a3846fa73 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/util.cpython-312.pyc differ diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/warping_network.cpython-312.pyc b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/warping_network.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..badb8cf1b4b29512e32a05a6450291d59b51324e Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/__pycache__/warping_network.cpython-312.pyc differ diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/appearance_feature_extractor.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/appearance_feature_extractor.py new file mode 100644 index 0000000000000000000000000000000000000000..9ec9b2809a66bb7a8d99ab876621af8db478fdbe --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/appearance_feature_extractor.py @@ -0,0 +1,48 @@ +# coding: utf-8 + +""" +Appearance extractor(F) defined in paper, which maps the source image s to a 3D appearance feature volume. +""" + +import torch +from torch import nn +from .util import SameBlock2d, DownBlock2d, ResBlock3d + + +class AppearanceFeatureExtractor(nn.Module): + + def __init__(self, image_channel, block_expansion, num_down_blocks, max_features, reshape_channel, reshape_depth, num_resblocks): + super(AppearanceFeatureExtractor, self).__init__() + self.image_channel = image_channel + self.block_expansion = block_expansion + self.num_down_blocks = num_down_blocks + self.max_features = max_features + self.reshape_channel = reshape_channel + self.reshape_depth = reshape_depth + + self.first = SameBlock2d(image_channel, block_expansion, kernel_size=(3, 3), padding=(1, 1)) + + down_blocks = [] + for i in range(num_down_blocks): + in_features = min(max_features, block_expansion * (2 ** i)) + out_features = min(max_features, block_expansion * (2 ** (i + 1))) + down_blocks.append(DownBlock2d(in_features, out_features, kernel_size=(3, 3), padding=(1, 1))) + self.down_blocks = nn.ModuleList(down_blocks) + + self.second = nn.Conv2d(in_channels=out_features, out_channels=max_features, kernel_size=1, stride=1) + + self.resblocks_3d = torch.nn.Sequential() + for i in range(num_resblocks): + self.resblocks_3d.add_module('3dr' + str(i), ResBlock3d(reshape_channel, kernel_size=3, padding=1)) + + def forward(self, source_image): + out = self.first(source_image) # Bx3x256x256 -> Bx64x256x256 + + for i in range(len(self.down_blocks)): + out = self.down_blocks[i](out) + out = self.second(out) + bs, c, h, w = out.shape # ->Bx512x64x64 + + f_s = out.view(bs, self.reshape_channel, self.reshape_depth, h, w) # ->Bx32x16x64x64 + f_s = self.resblocks_3d(f_s) # ->Bx32x16x64x64 + return f_s diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/convnextv2.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/convnextv2.py new file mode 100644 index 0000000000000000000000000000000000000000..1dd14cffa191ba9a1c95c118d730c06ffa45382a --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/convnextv2.py @@ -0,0 +1,149 @@ +# coding: utf-8 + +""" +This moudle is adapted to the ConvNeXtV2 version for the extraction of implicit keypoints, poses, and expression deformation. +""" + +import torch +import torch.nn as nn +# from timm.models.layers import trunc_normal_, DropPath +from .util import LayerNorm, DropPath, trunc_normal_, GRN + +__all__ = ['convnextv2_tiny'] + + +class Block(nn.Module): + """ ConvNeXtV2 Block. + + Args: + dim (int): Number of input channels. + drop_path (float): Stochastic depth rate. Default: 0.0 + """ + + def __init__(self, dim, drop_path=0.): + super().__init__() + self.dwconv = nn.Conv2d(dim, dim, kernel_size=7, padding=3, groups=dim) # depthwise conv + self.norm = LayerNorm(dim, eps=1e-6) + self.pwconv1 = nn.Linear(dim, 4 * dim) # pointwise/1x1 convs, implemented with linear layers + self.act = nn.GELU() + self.grn = GRN(4 * dim) + self.pwconv2 = nn.Linear(4 * dim, dim) + self.drop_path = DropPath(drop_path) if drop_path > 0. else nn.Identity() + + def forward(self, x): + input = x + x = self.dwconv(x) + x = x.permute(0, 2, 3, 1) # (N, C, H, W) -> (N, H, W, C) + x = self.norm(x) + x = self.pwconv1(x) + x = self.act(x) + x = self.grn(x) + x = self.pwconv2(x) + x = x.permute(0, 3, 1, 2) # (N, H, W, C) -> (N, C, H, W) + + x = input + self.drop_path(x) + return x + + +class ConvNeXtV2(nn.Module): + """ ConvNeXt V2 + + Args: + in_chans (int): Number of input image channels. Default: 3 + num_classes (int): Number of classes for classification head. Default: 1000 + depths (tuple(int)): Number of blocks at each stage. Default: [3, 3, 9, 3] + dims (int): Feature dimension at each stage. Default: [96, 192, 384, 768] + drop_path_rate (float): Stochastic depth rate. Default: 0. + head_init_scale (float): Init scaling value for classifier weights and biases. Default: 1. + """ + + def __init__( + self, + in_chans=3, + depths=[3, 3, 9, 3], + dims=[96, 192, 384, 768], + drop_path_rate=0., + **kwargs + ): + super().__init__() + self.depths = depths + self.downsample_layers = nn.ModuleList() # stem and 3 intermediate downsampling conv layers + stem = nn.Sequential( + nn.Conv2d(in_chans, dims[0], kernel_size=4, stride=4), + LayerNorm(dims[0], eps=1e-6, data_format="channels_first") + ) + self.downsample_layers.append(stem) + for i in range(3): + downsample_layer = nn.Sequential( + LayerNorm(dims[i], eps=1e-6, data_format="channels_first"), + nn.Conv2d(dims[i], dims[i+1], kernel_size=2, stride=2), + ) + self.downsample_layers.append(downsample_layer) + + self.stages = nn.ModuleList() # 4 feature resolution stages, each consisting of multiple residual blocks + dp_rates = [x.item() for x in torch.linspace(0, drop_path_rate, sum(depths))] + cur = 0 + for i in range(4): + stage = nn.Sequential( + *[Block(dim=dims[i], drop_path=dp_rates[cur + j]) for j in range(depths[i])] + ) + self.stages.append(stage) + cur += depths[i] + + self.norm = nn.LayerNorm(dims[-1], eps=1e-6) # final norm layer + + # NOTE: the output semantic items + num_bins = kwargs.get('num_bins', 66) + num_kp = kwargs.get('num_kp', 24) # the number of implicit keypoints + self.fc_kp = nn.Linear(dims[-1], 3 * num_kp) # implicit keypoints + + # print('dims[-1]: ', dims[-1]) + self.fc_scale = nn.Linear(dims[-1], 1) # scale + self.fc_pitch = nn.Linear(dims[-1], num_bins) # pitch bins + self.fc_yaw = nn.Linear(dims[-1], num_bins) # yaw bins + self.fc_roll = nn.Linear(dims[-1], num_bins) # roll bins + self.fc_t = nn.Linear(dims[-1], 3) # translation + self.fc_exp = nn.Linear(dims[-1], 3 * num_kp) # expression / delta + + def _init_weights(self, m): + if isinstance(m, (nn.Conv2d, nn.Linear)): + trunc_normal_(m.weight, std=.02) + nn.init.constant_(m.bias, 0) + + def forward_features(self, x): + for i in range(4): + x = self.downsample_layers[i](x) + x = self.stages[i](x) + return self.norm(x.mean([-2, -1])) # global average pooling, (N, C, H, W) -> (N, C) + + def forward(self, x): + x = self.forward_features(x) + + # implicit keypoints + kp = self.fc_kp(x) + + # pose and expression deformation + pitch = self.fc_pitch(x) + yaw = self.fc_yaw(x) + roll = self.fc_roll(x) + t = self.fc_t(x) + exp = self.fc_exp(x) + scale = self.fc_scale(x) + + ret_dct = { + 'pitch': pitch, + 'yaw': yaw, + 'roll': roll, + 't': t, + 'exp': exp, + 'scale': scale, + + 'kp': kp, # canonical keypoint + } + + return ret_dct + + +def convnextv2_tiny(**kwargs): + model = ConvNeXtV2(depths=[3, 3, 9, 3], dims=[96, 192, 384, 768], **kwargs) + return model diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/dense_motion.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/dense_motion.py new file mode 100644 index 0000000000000000000000000000000000000000..bb55206e6baac1fb285b6761f5a97352caee02df --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/dense_motion.py @@ -0,0 +1,104 @@ +# coding: utf-8 + +""" +The module that predicting a dense motion from sparse motion representation given by kp_source and kp_driving +""" + +from torch import nn +import torch.nn.functional as F +import torch +from .util import Hourglass, make_coordinate_grid, kp2gaussian + + +class DenseMotionNetwork(nn.Module): + def __init__(self, block_expansion, num_blocks, max_features, num_kp, feature_channel, reshape_depth, compress, estimate_occlusion_map=True): + super(DenseMotionNetwork, self).__init__() + self.hourglass = Hourglass(block_expansion=block_expansion, in_features=(num_kp+1)*(compress+1), max_features=max_features, num_blocks=num_blocks) # ~60+G + + self.mask = nn.Conv3d(self.hourglass.out_filters, num_kp + 1, kernel_size=7, padding=3) # 65G! NOTE: computation cost is large + self.compress = nn.Conv3d(feature_channel, compress, kernel_size=1) # 0.8G + self.norm = nn.BatchNorm3d(compress, affine=True) + self.num_kp = num_kp + self.flag_estimate_occlusion_map = estimate_occlusion_map + + if self.flag_estimate_occlusion_map: + self.occlusion = nn.Conv2d(self.hourglass.out_filters*reshape_depth, 1, kernel_size=7, padding=3) + else: + self.occlusion = None + + def create_sparse_motions(self, feature, kp_driving, kp_source): + bs, _, d, h, w = feature.shape # (bs, 4, 16, 64, 64) + identity_grid = make_coordinate_grid((d, h, w), ref=kp_source) # (16, 64, 64, 3) + identity_grid = identity_grid.view(1, 1, d, h, w, 3) # (1, 1, d=16, h=64, w=64, 3) + coordinate_grid = identity_grid - kp_driving.view(bs, self.num_kp, 1, 1, 1, 3) + + k = coordinate_grid.shape[1] + + # NOTE: there lacks an one-order flow + driving_to_source = coordinate_grid + kp_source.view(bs, self.num_kp, 1, 1, 1, 3) # (bs, num_kp, d, h, w, 3) + + # adding background feature + identity_grid = identity_grid.repeat(bs, 1, 1, 1, 1, 1) + sparse_motions = torch.cat([identity_grid, driving_to_source], dim=1) # (bs, 1+num_kp, d, h, w, 3) + return sparse_motions + + def create_deformed_feature(self, feature, sparse_motions): + bs, _, d, h, w = feature.shape + feature_repeat = feature.unsqueeze(1).unsqueeze(1).repeat(1, self.num_kp+1, 1, 1, 1, 1, 1) # (bs, num_kp+1, 1, c, d, h, w) + feature_repeat = feature_repeat.view(bs * (self.num_kp+1), -1, d, h, w) # (bs*(num_kp+1), c, d, h, w) + sparse_motions = sparse_motions.view((bs * (self.num_kp+1), d, h, w, -1)) # (bs*(num_kp+1), d, h, w, 3) + sparse_deformed = F.grid_sample(feature_repeat, sparse_motions, align_corners=False) + sparse_deformed = sparse_deformed.view((bs, self.num_kp+1, -1, d, h, w)) # (bs, num_kp+1, c, d, h, w) + + return sparse_deformed + + def create_heatmap_representations(self, feature, kp_driving, kp_source): + spatial_size = feature.shape[3:] # (d=16, h=64, w=64) + gaussian_driving = kp2gaussian(kp_driving, spatial_size=spatial_size, kp_variance=0.01) # (bs, num_kp, d, h, w) + gaussian_source = kp2gaussian(kp_source, spatial_size=spatial_size, kp_variance=0.01) # (bs, num_kp, d, h, w) + heatmap = gaussian_driving - gaussian_source # (bs, num_kp, d, h, w) + + # adding background feature + zeros = torch.zeros(heatmap.shape[0], 1, spatial_size[0], spatial_size[1], spatial_size[2]).type(heatmap.dtype).to(heatmap.device) + heatmap = torch.cat([zeros, heatmap], dim=1) + heatmap = heatmap.unsqueeze(2) # (bs, 1+num_kp, 1, d, h, w) + return heatmap + + def forward(self, feature, kp_driving, kp_source): + bs, _, d, h, w = feature.shape # (bs, 32, 16, 64, 64) + + feature = self.compress(feature) # (bs, 4, 16, 64, 64) + feature = self.norm(feature) # (bs, 4, 16, 64, 64) + feature = F.relu(feature) # (bs, 4, 16, 64, 64) + + out_dict = dict() + + # 1. deform 3d feature + sparse_motion = self.create_sparse_motions(feature, kp_driving, kp_source) # (bs, 1+num_kp, d, h, w, 3) + deformed_feature = self.create_deformed_feature(feature, sparse_motion) # (bs, 1+num_kp, c=4, d=16, h=64, w=64) + + # 2. (bs, 1+num_kp, d, h, w) + heatmap = self.create_heatmap_representations(deformed_feature, kp_driving, kp_source) # (bs, 1+num_kp, 1, d, h, w) + + input = torch.cat([heatmap, deformed_feature], dim=2) # (bs, 1+num_kp, c=5, d=16, h=64, w=64) + input = input.view(bs, -1, d, h, w) # (bs, (1+num_kp)*c=105, d=16, h=64, w=64) + + prediction = self.hourglass(input) + + mask = self.mask(prediction) + mask = F.softmax(mask, dim=1) # (bs, 1+num_kp, d=16, h=64, w=64) + out_dict['mask'] = mask + mask = mask.unsqueeze(2) # (bs, num_kp+1, 1, d, h, w) + sparse_motion = sparse_motion.permute(0, 1, 5, 2, 3, 4) # (bs, num_kp+1, 3, d, h, w) + deformation = (sparse_motion * mask).sum(dim=1) # (bs, 3, d, h, w) mask take effect in this place + deformation = deformation.permute(0, 2, 3, 4, 1) # (bs, d, h, w, 3) + + out_dict['deformation'] = deformation + + if self.flag_estimate_occlusion_map: + bs, _, d, h, w = prediction.shape + prediction_reshape = prediction.view(bs, -1, h, w) + occlusion_map = torch.sigmoid(self.occlusion(prediction_reshape)) # Bx1x64x64 + out_dict['occlusion_map'] = occlusion_map + + return out_dict diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/motion_extractor.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/motion_extractor.py new file mode 100644 index 0000000000000000000000000000000000000000..9318c101c507da2eae37b64c1bdbf34079c23d23 --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/motion_extractor.py @@ -0,0 +1,35 @@ +# coding: utf-8 + +""" +Motion extractor(M), which directly predicts the canonical keypoints, head pose and expression deformation of the input image +""" + +from torch import nn +import torch + +from .convnextv2 import convnextv2_tiny +from .util import filter_state_dict + +model_dict = { + 'convnextv2_tiny': convnextv2_tiny, +} + + +class MotionExtractor(nn.Module): + def __init__(self, **kwargs): + super(MotionExtractor, self).__init__() + + # default is convnextv2_base + backbone = kwargs.get('backbone', 'convnextv2_tiny') + self.detector = model_dict.get(backbone)(**kwargs) + + def load_pretrained(self, init_path: str): + if init_path not in (None, ''): + state_dict = torch.load(init_path, map_location=lambda storage, loc: storage)['model'] + state_dict = filter_state_dict(state_dict, remove_name='head') + ret = self.detector.load_state_dict(state_dict, strict=False) + print(f'Load pretrained model from {init_path}, ret: {ret}') + + def forward(self, x): + out = self.detector(x) + return out diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/spade_generator.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/spade_generator.py new file mode 100644 index 0000000000000000000000000000000000000000..314cd74d9d03785fe9b380b50edaad2081fcf192 --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/spade_generator.py @@ -0,0 +1,59 @@ +# coding: utf-8 + +""" +Spade decoder(G) defined in the paper, which input the warped feature to generate the animated image. +""" + +import torch +from torch import nn +import torch.nn.functional as F +from .util import SPADEResnetBlock + + +class SPADEDecoder(nn.Module): + def __init__(self, upscale=1, max_features=256, block_expansion=64, out_channels=64, num_down_blocks=2): + for i in range(num_down_blocks): + input_channels = min(max_features, block_expansion * (2 ** (i + 1))) + self.upscale = upscale + super().__init__() + norm_G = 'spadespectralinstance' + label_num_channels = input_channels # 256 + + self.fc = nn.Conv2d(input_channels, 2 * input_channels, 3, padding=1) + self.G_middle_0 = SPADEResnetBlock(2 * input_channels, 2 * input_channels, norm_G, label_num_channels) + self.G_middle_1 = SPADEResnetBlock(2 * input_channels, 2 * input_channels, norm_G, label_num_channels) + self.G_middle_2 = SPADEResnetBlock(2 * input_channels, 2 * input_channels, norm_G, label_num_channels) + self.G_middle_3 = SPADEResnetBlock(2 * input_channels, 2 * input_channels, norm_G, label_num_channels) + self.G_middle_4 = SPADEResnetBlock(2 * input_channels, 2 * input_channels, norm_G, label_num_channels) + self.G_middle_5 = SPADEResnetBlock(2 * input_channels, 2 * input_channels, norm_G, label_num_channels) + self.up_0 = SPADEResnetBlock(2 * input_channels, input_channels, norm_G, label_num_channels) + self.up_1 = SPADEResnetBlock(input_channels, out_channels, norm_G, label_num_channels) + self.up = nn.Upsample(scale_factor=2) + + if self.upscale is None or self.upscale <= 1: + self.conv_img = nn.Conv2d(out_channels, 3, 3, padding=1) + else: + self.conv_img = nn.Sequential( + nn.Conv2d(out_channels, 3 * (2 * 2), kernel_size=3, padding=1), + nn.PixelShuffle(upscale_factor=2) + ) + + def forward(self, feature): + seg = feature # Bx256x64x64 + x = self.fc(feature) # Bx512x64x64 + x = self.G_middle_0(x, seg) + x = self.G_middle_1(x, seg) + x = self.G_middle_2(x, seg) + x = self.G_middle_3(x, seg) + x = self.G_middle_4(x, seg) + x = self.G_middle_5(x, seg) + + x = self.up(x) # Bx512x64x64 -> Bx512x128x128 + x = self.up_0(x, seg) # Bx512x128x128 -> Bx256x128x128 + x = self.up(x) # Bx256x128x128 -> Bx256x256x256 + x = self.up_1(x, seg) # Bx256x256x256 -> Bx64x256x256 + + x = self.conv_img(F.leaky_relu(x, 2e-1)) # Bx64x256x256 -> Bx3xHxW + x = torch.sigmoid(x) # Bx3xHxW + + return x \ No newline at end of file diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/stitching_retargeting_network.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/stitching_retargeting_network.py new file mode 100644 index 0000000000000000000000000000000000000000..b459c4b3765dab58b2ff5e5325711679a350bee0 --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/stitching_retargeting_network.py @@ -0,0 +1,38 @@ +# coding: utf-8 + +""" +Stitching module(S) and two retargeting modules(R) defined in the paper. + +- The stitching module pastes the animated portrait back into the original image space without pixel misalignment, such as in +the stitching region. + +- The eyes retargeting module is designed to address the issue of incomplete eye closure during cross-id reenactment, especially +when a person with small eyes drives a person with larger eyes. + +- The lip retargeting module is designed similarly to the eye retargeting module, and can also normalize the input by ensuring that +the lips are in a closed state, which facilitates better animation driving. +""" +from torch import nn + + +class StitchingRetargetingNetwork(nn.Module): + def __init__(self, input_size, hidden_sizes, output_size): + super(StitchingRetargetingNetwork, self).__init__() + layers = [] + for i in range(len(hidden_sizes)): + if i == 0: + layers.append(nn.Linear(input_size, hidden_sizes[i])) + else: + layers.append(nn.Linear(hidden_sizes[i - 1], hidden_sizes[i])) + layers.append(nn.ReLU(inplace=True)) + layers.append(nn.Linear(hidden_sizes[-1], output_size)) + self.mlp = nn.Sequential(*layers) + + def initialize_weights_to_zero(self): + for m in self.modules(): + if isinstance(m, nn.Linear): + nn.init.zeros_(m.weight) + nn.init.zeros_(m.bias) + + def forward(self, x): + return self.mlp(x) diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/util.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/util.py new file mode 100644 index 0000000000000000000000000000000000000000..7dddb99b8085ce4f753e295d9db192a06a725baf --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/util.py @@ -0,0 +1,441 @@ +# coding: utf-8 + +""" +This file defines various neural network modules and utility functions, including convolutional and residual blocks, +normalizations, and functions for spatial transformation and tensor manipulation. +""" + +from torch import nn +import torch.nn.functional as F +import torch +import torch.nn.utils.spectral_norm as spectral_norm +import math +import warnings + + +def kp2gaussian(kp, spatial_size, kp_variance): + """ + Transform a keypoint into gaussian like representation + """ + mean = kp + + coordinate_grid = make_coordinate_grid(spatial_size, mean) + number_of_leading_dimensions = len(mean.shape) - 1 + shape = (1,) * number_of_leading_dimensions + coordinate_grid.shape + coordinate_grid = coordinate_grid.view(*shape) + repeats = mean.shape[:number_of_leading_dimensions] + (1, 1, 1, 1) + coordinate_grid = coordinate_grid.repeat(*repeats) + + # Preprocess kp shape + shape = mean.shape[:number_of_leading_dimensions] + (1, 1, 1, 3) + mean = mean.view(*shape) + + mean_sub = (coordinate_grid - mean) + + out = torch.exp(-0.5 * (mean_sub ** 2).sum(-1) / kp_variance) + + return out + + +def make_coordinate_grid(spatial_size, ref, **kwargs): + d, h, w = spatial_size + x = torch.arange(w).type(ref.dtype).to(ref.device) + y = torch.arange(h).type(ref.dtype).to(ref.device) + z = torch.arange(d).type(ref.dtype).to(ref.device) + + # NOTE: must be right-down-in + x = (2 * (x / (w - 1)) - 1) # the x axis faces to the right + y = (2 * (y / (h - 1)) - 1) # the y axis faces to the bottom + z = (2 * (z / (d - 1)) - 1) # the z axis faces to the inner + + yy = y.view(1, -1, 1).repeat(d, 1, w) + xx = x.view(1, 1, -1).repeat(d, h, 1) + zz = z.view(-1, 1, 1).repeat(1, h, w) + + meshed = torch.cat([xx.unsqueeze_(3), yy.unsqueeze_(3), zz.unsqueeze_(3)], 3) + + return meshed + + +class ConvT2d(nn.Module): + """ + Upsampling block for use in decoder. + """ + + def __init__(self, in_features, out_features, kernel_size=3, stride=2, padding=1, output_padding=1): + super(ConvT2d, self).__init__() + + self.convT = nn.ConvTranspose2d(in_features, out_features, kernel_size=kernel_size, stride=stride, + padding=padding, output_padding=output_padding) + self.norm = nn.InstanceNorm2d(out_features) + + def forward(self, x): + out = self.convT(x) + out = self.norm(out) + out = F.leaky_relu(out) + return out + + +class ResBlock3d(nn.Module): + """ + Res block, preserve spatial resolution. + """ + + def __init__(self, in_features, kernel_size, padding): + super(ResBlock3d, self).__init__() + self.conv1 = nn.Conv3d(in_channels=in_features, out_channels=in_features, kernel_size=kernel_size, padding=padding) + self.conv2 = nn.Conv3d(in_channels=in_features, out_channels=in_features, kernel_size=kernel_size, padding=padding) + self.norm1 = nn.BatchNorm3d(in_features, affine=True) + self.norm2 = nn.BatchNorm3d(in_features, affine=True) + + def forward(self, x): + out = self.norm1(x) + out = F.relu(out) + out = self.conv1(out) + out = self.norm2(out) + out = F.relu(out) + out = self.conv2(out) + out += x + return out + + +class UpBlock3d(nn.Module): + """ + Upsampling block for use in decoder. + """ + + def __init__(self, in_features, out_features, kernel_size=3, padding=1, groups=1): + super(UpBlock3d, self).__init__() + + self.conv = nn.Conv3d(in_channels=in_features, out_channels=out_features, kernel_size=kernel_size, + padding=padding, groups=groups) + self.norm = nn.BatchNorm3d(out_features, affine=True) + + def forward(self, x): + out = F.interpolate(x, scale_factor=(1, 2, 2)) + out = self.conv(out) + out = self.norm(out) + out = F.relu(out) + return out + + +class DownBlock2d(nn.Module): + """ + Downsampling block for use in encoder. + """ + + def __init__(self, in_features, out_features, kernel_size=3, padding=1, groups=1): + super(DownBlock2d, self).__init__() + self.conv = nn.Conv2d(in_channels=in_features, out_channels=out_features, kernel_size=kernel_size, padding=padding, groups=groups) + self.norm = nn.BatchNorm2d(out_features, affine=True) + self.pool = nn.AvgPool2d(kernel_size=(2, 2)) + + def forward(self, x): + out = self.conv(x) + out = self.norm(out) + out = F.relu(out) + out = self.pool(out) + return out + + +class DownBlock3d(nn.Module): + """ + Downsampling block for use in encoder. + """ + + def __init__(self, in_features, out_features, kernel_size=3, padding=1, groups=1): + super(DownBlock3d, self).__init__() + ''' + self.conv = nn.Conv3d(in_channels=in_features, out_channels=out_features, kernel_size=kernel_size, + padding=padding, groups=groups, stride=(1, 2, 2)) + ''' + self.conv = nn.Conv3d(in_channels=in_features, out_channels=out_features, kernel_size=kernel_size, + padding=padding, groups=groups) + self.norm = nn.BatchNorm3d(out_features, affine=True) + self.pool = nn.AvgPool3d(kernel_size=(1, 2, 2)) + + def forward(self, x): + out = self.conv(x) + out = self.norm(out) + out = F.relu(out) + out = self.pool(out) + return out + + +class SameBlock2d(nn.Module): + """ + Simple block, preserve spatial resolution. + """ + + def __init__(self, in_features, out_features, groups=1, kernel_size=3, padding=1, lrelu=False): + super(SameBlock2d, self).__init__() + self.conv = nn.Conv2d(in_channels=in_features, out_channels=out_features, kernel_size=kernel_size, padding=padding, groups=groups) + self.norm = nn.BatchNorm2d(out_features, affine=True) + if lrelu: + self.ac = nn.LeakyReLU() + else: + self.ac = nn.ReLU() + + def forward(self, x): + out = self.conv(x) + out = self.norm(out) + out = self.ac(out) + return out + + +class Encoder(nn.Module): + """ + Hourglass Encoder + """ + + def __init__(self, block_expansion, in_features, num_blocks=3, max_features=256): + super(Encoder, self).__init__() + + down_blocks = [] + for i in range(num_blocks): + down_blocks.append(DownBlock3d(in_features if i == 0 else min(max_features, block_expansion * (2 ** i)), min(max_features, block_expansion * (2 ** (i + 1))), kernel_size=3, padding=1)) + self.down_blocks = nn.ModuleList(down_blocks) + + def forward(self, x): + outs = [x] + for down_block in self.down_blocks: + outs.append(down_block(outs[-1])) + return outs + + +class Decoder(nn.Module): + """ + Hourglass Decoder + """ + + def __init__(self, block_expansion, in_features, num_blocks=3, max_features=256): + super(Decoder, self).__init__() + + up_blocks = [] + + for i in range(num_blocks)[::-1]: + in_filters = (1 if i == num_blocks - 1 else 2) * min(max_features, block_expansion * (2 ** (i + 1))) + out_filters = min(max_features, block_expansion * (2 ** i)) + up_blocks.append(UpBlock3d(in_filters, out_filters, kernel_size=3, padding=1)) + + self.up_blocks = nn.ModuleList(up_blocks) + self.out_filters = block_expansion + in_features + + self.conv = nn.Conv3d(in_channels=self.out_filters, out_channels=self.out_filters, kernel_size=3, padding=1) + self.norm = nn.BatchNorm3d(self.out_filters, affine=True) + + def forward(self, x): + out = x.pop() + for up_block in self.up_blocks: + out = up_block(out) + skip = x.pop() + out = torch.cat([out, skip], dim=1) + out = self.conv(out) + out = self.norm(out) + out = F.relu(out) + return out + + +class Hourglass(nn.Module): + """ + Hourglass architecture. + """ + + def __init__(self, block_expansion, in_features, num_blocks=3, max_features=256): + super(Hourglass, self).__init__() + self.encoder = Encoder(block_expansion, in_features, num_blocks, max_features) + self.decoder = Decoder(block_expansion, in_features, num_blocks, max_features) + self.out_filters = self.decoder.out_filters + + def forward(self, x): + return self.decoder(self.encoder(x)) + + +class SPADE(nn.Module): + def __init__(self, norm_nc, label_nc): + super().__init__() + + self.param_free_norm = nn.InstanceNorm2d(norm_nc, affine=False) + nhidden = 128 + + self.mlp_shared = nn.Sequential( + nn.Conv2d(label_nc, nhidden, kernel_size=3, padding=1), + nn.ReLU()) + self.mlp_gamma = nn.Conv2d(nhidden, norm_nc, kernel_size=3, padding=1) + self.mlp_beta = nn.Conv2d(nhidden, norm_nc, kernel_size=3, padding=1) + + def forward(self, x, segmap): + normalized = self.param_free_norm(x) + segmap = F.interpolate(segmap, size=x.size()[2:], mode='nearest') + actv = self.mlp_shared(segmap) + gamma = self.mlp_gamma(actv) + beta = self.mlp_beta(actv) + out = normalized * (1 + gamma) + beta + return out + + +class SPADEResnetBlock(nn.Module): + def __init__(self, fin, fout, norm_G, label_nc, use_se=False, dilation=1): + super().__init__() + # Attributes + self.learned_shortcut = (fin != fout) + fmiddle = min(fin, fout) + self.use_se = use_se + # create conv layers + self.conv_0 = nn.Conv2d(fin, fmiddle, kernel_size=3, padding=dilation, dilation=dilation) + self.conv_1 = nn.Conv2d(fmiddle, fout, kernel_size=3, padding=dilation, dilation=dilation) + if self.learned_shortcut: + self.conv_s = nn.Conv2d(fin, fout, kernel_size=1, bias=False) + # apply spectral norm if specified + if 'spectral' in norm_G: + self.conv_0 = spectral_norm(self.conv_0) + self.conv_1 = spectral_norm(self.conv_1) + if self.learned_shortcut: + self.conv_s = spectral_norm(self.conv_s) + # define normalization layers + self.norm_0 = SPADE(fin, label_nc) + self.norm_1 = SPADE(fmiddle, label_nc) + if self.learned_shortcut: + self.norm_s = SPADE(fin, label_nc) + + def forward(self, x, seg1): + x_s = self.shortcut(x, seg1) + dx = self.conv_0(self.actvn(self.norm_0(x, seg1))) + dx = self.conv_1(self.actvn(self.norm_1(dx, seg1))) + out = x_s + dx + return out + + def shortcut(self, x, seg1): + if self.learned_shortcut: + x_s = self.conv_s(self.norm_s(x, seg1)) + else: + x_s = x + return x_s + + def actvn(self, x): + return F.leaky_relu(x, 2e-1) + + +def filter_state_dict(state_dict, remove_name='fc'): + new_state_dict = {} + for key in state_dict: + if remove_name in key: + continue + new_state_dict[key] = state_dict[key] + return new_state_dict + + +class GRN(nn.Module): + """ GRN (Global Response Normalization) layer + """ + + def __init__(self, dim): + super().__init__() + self.gamma = nn.Parameter(torch.zeros(1, 1, 1, dim)) + self.beta = nn.Parameter(torch.zeros(1, 1, 1, dim)) + + def forward(self, x): + Gx = torch.norm(x, p=2, dim=(1, 2), keepdim=True) + Nx = Gx / (Gx.mean(dim=-1, keepdim=True) + 1e-6) + return self.gamma * (x * Nx) + self.beta + x + + +class LayerNorm(nn.Module): + r""" LayerNorm that supports two data formats: channels_last (default) or channels_first. + The ordering of the dimensions in the inputs. channels_last corresponds to inputs with + shape (batch_size, height, width, channels) while channels_first corresponds to inputs + with shape (batch_size, channels, height, width). + """ + + def __init__(self, normalized_shape, eps=1e-6, data_format="channels_last"): + super().__init__() + self.weight = nn.Parameter(torch.ones(normalized_shape)) + self.bias = nn.Parameter(torch.zeros(normalized_shape)) + self.eps = eps + self.data_format = data_format + if self.data_format not in ["channels_last", "channels_first"]: + raise NotImplementedError + self.normalized_shape = (normalized_shape, ) + + def forward(self, x): + if self.data_format == "channels_last": + return F.layer_norm(x, self.normalized_shape, self.weight, self.bias, self.eps) + elif self.data_format == "channels_first": + u = x.mean(1, keepdim=True) + s = (x - u).pow(2).mean(1, keepdim=True) + x = (x - u) / torch.sqrt(s + self.eps) + x = self.weight[:, None, None] * x + self.bias[:, None, None] + return x + + +def _no_grad_trunc_normal_(tensor, mean, std, a, b): + # Cut & paste from PyTorch official master until it's in a few official releases - RW + # Method based on https://people.sc.fsu.edu/~jburkardt/presentations/truncated_normal.pdf + def norm_cdf(x): + # Computes standard normal cumulative distribution function + return (1. + math.erf(x / math.sqrt(2.))) / 2. + + if (mean < a - 2 * std) or (mean > b + 2 * std): + warnings.warn("mean is more than 2 std from [a, b] in nn.init.trunc_normal_. " + "The distribution of values may be incorrect.", + stacklevel=2) + + with torch.no_grad(): + # Values are generated by using a truncated uniform distribution and + # then using the inverse CDF for the normal distribution. + # Get upper and lower cdf values + l = norm_cdf((a - mean) / std) + u = norm_cdf((b - mean) / std) + + # Uniformly fill tensor with values from [l, u], then translate to + # [2l-1, 2u-1]. + tensor.uniform_(2 * l - 1, 2 * u - 1) + + # Use inverse cdf transform for normal distribution to get truncated + # standard normal + tensor.erfinv_() + + # Transform to proper mean, std + tensor.mul_(std * math.sqrt(2.)) + tensor.add_(mean) + + # Clamp to ensure it's in the proper range + tensor.clamp_(min=a, max=b) + return tensor + + +def drop_path(x, drop_prob=0., training=False, scale_by_keep=True): + """ Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks). + + This is the same as the DropConnect impl I created for EfficientNet, etc networks, however, + the original name is misleading as 'Drop Connect' is a different form of dropout in a separate paper... + See discussion: https://github.com/tensorflow/tpu/issues/494#issuecomment-532968956 ... I've opted for + changing the layer and argument names to 'drop path' rather than mix DropConnect as a layer name and use + 'survival rate' as the argument. + + """ + if drop_prob == 0. or not training: + return x + keep_prob = 1 - drop_prob + shape = (x.shape[0],) + (1,) * (x.ndim - 1) # work with diff dim tensors, not just 2D ConvNets + random_tensor = x.new_empty(shape).bernoulli_(keep_prob) + if keep_prob > 0.0 and scale_by_keep: + random_tensor.div_(keep_prob) + return x * random_tensor + + +class DropPath(nn.Module): + """ Drop paths (Stochastic Depth) per sample (when applied in main path of residual blocks). + """ + + def __init__(self, drop_prob=None, scale_by_keep=True): + super(DropPath, self).__init__() + self.drop_prob = drop_prob + self.scale_by_keep = scale_by_keep + + def forward(self, x): + return drop_path(x, self.drop_prob, self.training, self.scale_by_keep) + + +def trunc_normal_(tensor, mean=0., std=1., a=-2., b=2.): + return _no_grad_trunc_normal_(tensor, mean, std, a, b) diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/warping_network.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/warping_network.py new file mode 100644 index 0000000000000000000000000000000000000000..64110dd963381d4c07c3925bdd9acdec73fe6a8c --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/LivePortrait/modules/warping_network.py @@ -0,0 +1,77 @@ +# coding: utf-8 + +""" +Warping field estimator(W) defined in the paper, which generates a warping field using the implicit +keypoint representations x_s and x_d, and employs this flow field to warp the source feature volume f_s. +""" + +from torch import nn +import torch.nn.functional as F +from .util import SameBlock2d +from .dense_motion import DenseMotionNetwork + + +class WarpingNetwork(nn.Module): + def __init__( + self, + num_kp, + block_expansion, + max_features, + num_down_blocks, + reshape_channel, + estimate_occlusion_map=False, + dense_motion_params=None, + **kwargs + ): + super(WarpingNetwork, self).__init__() + + self.upscale = kwargs.get('upscale', 1) + self.flag_use_occlusion_map = kwargs.get('flag_use_occlusion_map', True) + + if dense_motion_params is not None: + self.dense_motion_network = DenseMotionNetwork( + num_kp=num_kp, + feature_channel=reshape_channel, + estimate_occlusion_map=estimate_occlusion_map, + **dense_motion_params + ) + else: + self.dense_motion_network = None + + self.third = SameBlock2d(max_features, block_expansion * (2 ** num_down_blocks), kernel_size=(3, 3), padding=(1, 1), lrelu=True) + self.fourth = nn.Conv2d(in_channels=block_expansion * (2 ** num_down_blocks), out_channels=block_expansion * (2 ** num_down_blocks), kernel_size=1, stride=1) + + self.estimate_occlusion_map = estimate_occlusion_map + + def deform_input(self, inp, deformation): + return F.grid_sample(inp, deformation, align_corners=False) + + def forward(self, feature_3d, kp_driving, kp_source): + if self.dense_motion_network is not None: + # Feature warper, Transforming feature representation according to deformation and occlusion + dense_motion = self.dense_motion_network( + feature=feature_3d, kp_driving=kp_driving, kp_source=kp_source + ) + if 'occlusion_map' in dense_motion: + occlusion_map = dense_motion['occlusion_map'] # Bx1x64x64 + else: + occlusion_map = None + + deformation = dense_motion['deformation'] # Bx16x64x64x3 + out = self.deform_input(feature_3d, deformation) # Bx32x16x64x64 + + bs, c, d, h, w = out.shape # Bx32x16x64x64 + out = out.view(bs, c * d, h, w) # -> Bx512x64x64 + out = self.third(out) # -> Bx256x64x64 + out = self.fourth(out) # -> Bx256x64x64 + + if self.flag_use_occlusion_map and (occlusion_map is not None): + out = out * occlusion_map + + ret_dct = { + 'occlusion_map': occlusion_map, + 'deformation': deformation, + 'out': out, + } + + return ret_dct diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/__init__.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/__pycache__/__init__.cpython-312.pyc b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/__pycache__/__init__.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..fab42adf52aa3c0eb6b8b8c345ae810dce1cea1c Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/__pycache__/__init__.cpython-312.pyc differ diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/__pycache__/camera.cpython-312.pyc b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/__pycache__/camera.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..75f9dc22efd8d055b5e9443751e755efbd318917 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/__pycache__/camera.cpython-312.pyc differ diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/__pycache__/helper.cpython-312.pyc b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/__pycache__/helper.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..1b7615f1b9e3df740cc6d2741677180af510eeab Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/__pycache__/helper.cpython-312.pyc differ diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/__pycache__/rprint.cpython-312.pyc b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/__pycache__/rprint.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..cfd1b1a4b44e66319037e662eed489c8d3ee81ea Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/__pycache__/rprint.cpython-312.pyc differ diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/camera.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/camera.py new file mode 100644 index 0000000000000000000000000000000000000000..79000f59143b625c2ab65a5ae80d2a0aa4201faa --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/camera.py @@ -0,0 +1,75 @@ +# coding: utf-8 + +""" +functions for processing and transforming 3D facial keypoints +""" + +import numpy as np +import torch +import torch.nn.functional as F + +PI = np.pi + + +def headpose_pred_to_degree(pred): + """ + pred: (bs, 66) or (bs, 1) or others + """ + if pred.ndim > 1 and pred.shape[1] == 66: + # NOTE: note that the average is modified to 97.5 + device = pred.device + idx_tensor = [idx for idx in range(0, 66)] + idx_tensor = torch.FloatTensor(idx_tensor).to(device) + pred = F.softmax(pred, dim=1) + degree = torch.sum(pred*idx_tensor, axis=1) * 3 - 97.5 + + return degree + + return pred + + +def get_rotation_matrix(pitch_, yaw_, roll_): + """ the input is in degree + """ + # calculate the rotation matrix: vps @ rot + + # transform to radian + pitch = pitch_ / 180 * PI + yaw = yaw_ / 180 * PI + roll = roll_ / 180 * PI + + device = pitch.device + + if pitch.ndim == 1: + pitch = pitch.unsqueeze(1) + if yaw.ndim == 1: + yaw = yaw.unsqueeze(1) + if roll.ndim == 1: + roll = roll.unsqueeze(1) + + # calculate the euler matrix + bs = pitch.shape[0] + ones = torch.ones([bs, 1]).to(device) + zeros = torch.zeros([bs, 1]).to(device) + x, y, z = pitch, yaw, roll + + rot_x = torch.cat([ + ones, zeros, zeros, + zeros, torch.cos(x), -torch.sin(x), + zeros, torch.sin(x), torch.cos(x) + ], dim=1).reshape([bs, 3, 3]) + + rot_y = torch.cat([ + torch.cos(y), zeros, torch.sin(y), + zeros, ones, zeros, + -torch.sin(y), zeros, torch.cos(y) + ], dim=1).reshape([bs, 3, 3]) + + rot_z = torch.cat([ + torch.cos(z), -torch.sin(z), zeros, + torch.sin(z), torch.cos(z), zeros, + zeros, zeros, ones + ], dim=1).reshape([bs, 3, 3]) + + rot = rot_z @ rot_y @ rot_x + return rot.permute(0, 2, 1) # transpose diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/face_analysis_diy.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/face_analysis_diy.py new file mode 100644 index 0000000000000000000000000000000000000000..b2175cba1544d01ab76f031444e560748af5f222 --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/face_analysis_diy.py @@ -0,0 +1,78 @@ +# coding: utf-8 + +""" +face detectoin and alignment using InsightFace +""" + +import numpy as np +from .rprint import rlog as log +from insightface.app import FaceAnalysis +from insightface.app.common import Face +from .timer import Timer + + +def sort_by_direction(faces, direction: str = 'large-small', face_center=None): + if len(faces) <= 0: + return faces + if direction == 'left-right': + return sorted(faces, key=lambda face: face['bbox'][0]) + if direction == 'right-left': + return sorted(faces, key=lambda face: face['bbox'][0], reverse=True) + if direction == 'top-bottom': + return sorted(faces, key=lambda face: face['bbox'][1]) + if direction == 'bottom-top': + return sorted(faces, key=lambda face: face['bbox'][1], reverse=True) + if direction == 'small-large': + return sorted(faces, key=lambda face: (face['bbox'][2] - face['bbox'][0]) * (face['bbox'][3] - face['bbox'][1])) + if direction == 'large-small': + return sorted(faces, key=lambda face: (face['bbox'][2] - face['bbox'][0]) * (face['bbox'][3] - face['bbox'][1]), reverse=True) + if direction == 'distance-from-retarget-face': + return sorted(faces, key=lambda face: (((face['bbox'][2]+face['bbox'][0])/2-face_center[0])**2+((face['bbox'][3]+face['bbox'][1])/2-face_center[1])**2)**0.5) + return faces + + +class FaceAnalysisDIY(FaceAnalysis): + def __init__(self, name='buffalo_l', root='~/.insightface', allowed_modules=None, **kwargs): + super().__init__(name=name, root=root, allowed_modules=allowed_modules, **kwargs) + + self.timer = Timer() + + def get(self, img_bgr, **kwargs): + max_num = kwargs.get('max_num', 0) # the number of the detected faces, 0 means no limit + flag_do_landmark_2d_106 = kwargs.get('flag_do_landmark_2d_106', True) # whether to do 106-point detection + direction = kwargs.get('direction', 'large-small') # sorting direction + face_center = None + + bboxes, kpss = self.det_model.detect(img_bgr, max_num=max_num, metric='default') + if bboxes.shape[0] == 0: + return [] + ret = [] + for i in range(bboxes.shape[0]): + bbox = bboxes[i, 0:4] + det_score = bboxes[i, 4] + kps = None + if kpss is not None: + kps = kpss[i] + face = Face(bbox=bbox, kps=kps, det_score=det_score) + for taskname, model in self.models.items(): + if taskname == 'detection': + continue + + if (not flag_do_landmark_2d_106) and taskname == 'landmark_2d_106': + continue + + # print(f'taskname: {taskname}') + model.get(img_bgr, face) + ret.append(face) + + ret = sort_by_direction(ret, direction, face_center) + return ret + + def warmup(self): + self.timer.tic() + + img_bgr = np.zeros((512, 512, 3), dtype=np.uint8) + self.get(img_bgr) + + elapse = self.timer.toc() + log(f'FaceAnalysisDIY warmup time: {elapse:.3f}s') diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/helper.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/helper.py new file mode 100644 index 0000000000000000000000000000000000000000..063c02a0765aa09b1e2a8942910b8a42e4748e04 --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/helper.py @@ -0,0 +1,124 @@ +# coding: utf-8 + +""" +utility functions and classes to handle feature extraction and model loading +""" + +import os +import os.path as osp +import cv2 +import torch +from rich.console import Console +from collections import OrderedDict + +from ..modules.spade_generator import SPADEDecoder +from ..modules.warping_network import WarpingNetwork +from ..modules.motion_extractor import MotionExtractor +from ..modules.appearance_feature_extractor import AppearanceFeatureExtractor +from ..modules.stitching_retargeting_network import StitchingRetargetingNetwork +from .rprint import rlog as log + + +def suffix(filename): + """a.jpg -> jpg""" + pos = filename.rfind(".") + if pos == -1: + return "" + return filename[pos + 1:] + + +def prefix(filename): + """a.jpg -> a""" + pos = filename.rfind(".") + if pos == -1: + return filename + return filename[:pos] + + +def basename(filename): + """a/b/c.jpg -> c""" + return prefix(osp.basename(filename)) + + +def is_video(file_path): + if file_path.lower().endswith((".mp4", ".mov", ".avi", ".webm")) or osp.isdir(file_path): + return True + return False + +def is_template(file_path): + if file_path.endswith(".pkl"): + return True + return False + + +def mkdir(d, log=False): + # return self-assined `d`, for one line code + if not osp.exists(d): + os.makedirs(d, exist_ok=True) + if log: + print(f"Make dir: {d}") + return d + + +def squeeze_tensor_to_numpy(tensor): + out = tensor.data.squeeze(0).cpu().numpy() + return out + + +def dct2cuda(dct: dict, device_id: int): + for key in dct: + dct[key] = torch.tensor(dct[key]).cuda(device_id) + return dct + + +def concat_feat(kp_source: torch.Tensor, kp_driving: torch.Tensor) -> torch.Tensor: + """ + kp_source: (bs, k, 3) + kp_driving: (bs, k, 3) + Return: (bs, 2k*3) + """ + bs_src = kp_source.shape[0] + bs_dri = kp_driving.shape[0] + assert bs_src == bs_dri, 'batch size must be equal' + + feat = torch.cat([kp_source.view(bs_src, -1), kp_driving.view(bs_dri, -1)], dim=1) + return feat + + +# get coefficients of Eqn. 7 +def calculate_transformation(config, s_kp_info, t_0_kp_info, t_i_kp_info, R_s, R_t_0, R_t_i): + if config.relative: + new_rotation = (R_t_i @ R_t_0.permute(0, 2, 1)) @ R_s + new_expression = s_kp_info['exp'] + (t_i_kp_info['exp'] - t_0_kp_info['exp']) + else: + new_rotation = R_t_i + new_expression = t_i_kp_info['exp'] + new_translation = s_kp_info['t'] + (t_i_kp_info['t'] - t_0_kp_info['t']) + new_translation[..., 2].fill_(0) # Keep the z-axis unchanged + new_scale = s_kp_info['scale'] * (t_i_kp_info['scale'] / t_0_kp_info['scale']) + return new_rotation, new_expression, new_translation, new_scale + +def load_description(fp): + with open(fp, 'r', encoding='utf-8') as f: + content = f.read() + return content + + +def resize_to_limit(img, max_dim=1280, n=2): + h, w = img.shape[:2] + if max_dim > 0 and max(h, w) > max_dim: + if h > w: + new_h = max_dim + new_w = int(w * (max_dim / h)) + else: + new_w = max_dim + new_h = int(h * (max_dim / w)) + img = cv2.resize(img, (new_w, new_h)) + n = max(n, 1) + new_h = img.shape[0] - (img.shape[0] % n) + new_w = img.shape[1] - (img.shape[1] % n) + if new_h == 0 or new_w == 0: + return img + if new_h != img.shape[0] or new_w != img.shape[1]: + img = img[:new_h, :new_w] + return img diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/io.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/io.py new file mode 100644 index 0000000000000000000000000000000000000000..e19065ca4bf59ef0586c7601ac80dd9ece927e6f --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/io.py @@ -0,0 +1,97 @@ +# coding: utf-8 + +import os +from glob import glob +import os.path as osp +import imageio +import numpy as np +import cv2; cv2.setNumThreads(0); cv2.ocl.setUseOpenCL(False) + + +def load_image_rgb(image_path: str): + if not osp.exists(image_path): + raise FileNotFoundError(f"Image not found: {image_path}") + img = cv2.imread(image_path, cv2.IMREAD_COLOR) + return cv2.cvtColor(img, cv2.COLOR_BGR2RGB) + + +def load_driving_info(driving_info): + driving_video_ori = [] + + def load_images_from_directory(directory): + image_paths = sorted(glob(osp.join(directory, '*.png')) + glob(osp.join(directory, '*.jpg'))) + return [load_image_rgb(im_path) for im_path in image_paths] + + def load_images_from_video(file_path): + reader = imageio.get_reader(file_path) + return [image for idx, image in enumerate(reader)] + + if osp.isdir(driving_info): + driving_video_ori = load_images_from_directory(driving_info) + elif osp.isfile(driving_info): + driving_video_ori = load_images_from_video(driving_info) + + return driving_video_ori + + +def contiguous(obj): + if not obj.flags.c_contiguous: + obj = obj.copy(order="C") + return obj + + +def _resize_to_limit(img: np.ndarray, max_dim=1920, n=2): + """ + ajust the size of the image so that the maximum dimension does not exceed max_dim, and the width and the height of the image are multiples of n. + :param img: the image to be processed. + :param max_dim: the maximum dimension constraint. + :param n: the number that needs to be multiples of. + :return: the adjusted image. + """ + h, w = img.shape[:2] + + # ajust the size of the image according to the maximum dimension + if max_dim > 0 and max(h, w) > max_dim: + if h > w: + new_h = max_dim + new_w = int(w * (max_dim / h)) + else: + new_w = max_dim + new_h = int(h * (max_dim / w)) + img = cv2.resize(img, (new_w, new_h)) + + # ensure that the image dimensions are multiples of n + n = max(n, 1) + new_h = img.shape[0] - (img.shape[0] % n) + new_w = img.shape[1] - (img.shape[1] % n) + + if new_h == 0 or new_w == 0: + # when the width or height is less than n, no need to process + return img + + if new_h != img.shape[0] or new_w != img.shape[1]: + img = img[:new_h, :new_w] + + return img + + +def load_img_online(obj, mode="bgr", **kwargs): + max_dim = kwargs.get("max_dim", 1920) + n = kwargs.get("n", 2) + if isinstance(obj, str): + if mode.lower() == "gray": + img = cv2.imread(obj, cv2.IMREAD_GRAYSCALE) + else: + img = cv2.imread(obj, cv2.IMREAD_COLOR) + else: + img = obj + + # Resize image to satisfy constraints + img = _resize_to_limit(img, max_dim=max_dim, n=n) + + if mode.lower() == "bgr": + return contiguous(img) + elif mode.lower() == "rgb": + return contiguous(img[..., ::-1]) + else: + raise Exception(f"Unknown mode {mode}") diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/resources/mask_template.png b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/resources/mask_template.png new file mode 100644 index 0000000000000000000000000000000000000000..bca6ca5977ba820d0d2c05b3793c6231cc82e715 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/resources/mask_template.png differ diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/rprint.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/rprint.py new file mode 100644 index 0000000000000000000000000000000000000000..18614e86d3b7e11552dd4f82b92712d56324ff07 --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/rprint.py @@ -0,0 +1,16 @@ +# coding: utf-8 + +""" +custom print and log functions +""" + +__all__ = ['rprint', 'rlog'] + +try: + from rich.console import Console + console = Console() + rprint = console.print + rlog = console.log +except: + rprint = print + rlog = print diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/timer.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/timer.py new file mode 100644 index 0000000000000000000000000000000000000000..792af243ad59ee7eb7768b40a7370c9d68bcff7c --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/timer.py @@ -0,0 +1,29 @@ +# coding: utf-8 + +""" +tools to measure elapsed time +""" + +import time + +class Timer(object): + """A simple timer.""" + + def __init__(self): + self.total_time = 0. + self.calls = 0 + self.start_time = 0. + self.diff = 0. + + def tic(self): + # using time.time instead of time.clock because time time.clock + # does not normalize for multithreading + self.start_time = time.time() + + def toc(self, average=True): + self.diff = time.time() - self.start_time + return self.diff + + def clear(self): + self.start_time = 0. + self.diff = 0. diff --git a/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/video.py b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/video.py new file mode 100644 index 0000000000000000000000000000000000000000..c1860eddcf585fa6a98fce51b838ed03959c9176 --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/LivePortrait/utils/video.py @@ -0,0 +1,142 @@ +# coding: utf-8 + +""" +functions for processing video +""" + +import os.path as osp +import numpy as np +import subprocess +import imageio +import cv2 + +# from rich.progress import track +from .helper import prefix +from .rprint import rprint as print + + +def exec_cmd(cmd): + subprocess.run(cmd, shell=True, check=True, stdout=subprocess.PIPE, stderr=subprocess.STDOUT) + + +def images2video(images, wfp, **kwargs): + fps = kwargs.get('fps', 30) + video_format = kwargs.get('format', 'mp4') # default is mp4 format + codec = kwargs.get('codec', 'libx264') # default is libx264 encoding + quality = kwargs.get('quality') # video quality + pixelformat = kwargs.get('pixelformat', 'yuv420p') # video pixel format + image_mode = kwargs.get('image_mode', 'rgb') + macro_block_size = kwargs.get('macro_block_size', 2) + ffmpeg_params = ['-crf', str(kwargs.get('crf', 18))] + + writer = imageio.get_writer( + wfp, fps=fps, format=video_format, + codec=codec, quality=quality, ffmpeg_params=ffmpeg_params, pixelformat=pixelformat, macro_block_size=macro_block_size + ) + + n = len(images) + print('writing',n) + for i in range(n): + if image_mode.lower() == 'bgr': + writer.append_data(images[i][..., ::-1]) + else: + writer.append_data(images[i]) + + writer.close() + + # print(f':smiley: Dump to {wfp}\n', style="bold green") + print(f'Dump to {wfp}\n') + + +def video2gif(video_fp, fps=30, size=256): + if osp.exists(video_fp): + d = osp.split(video_fp)[0] + fn = prefix(osp.basename(video_fp)) + palette_wfp = osp.join(d, 'palette.png') + gif_wfp = osp.join(d, f'{fn}.gif') + # generate the palette + cmd = f'ffmpeg -i {video_fp} -vf "fps={fps},scale={size}:-1:flags=lanczos,palettegen" {palette_wfp} -y' + exec_cmd(cmd) + # use the palette to generate the gif + cmd = f'ffmpeg -i {video_fp} -i {palette_wfp} -filter_complex "fps={fps},scale={size}:-1:flags=lanczos[x];[x][1:v]paletteuse" {gif_wfp} -y' + exec_cmd(cmd) + else: + print(f'video_fp: {video_fp} not exists!') + + +def merge_audio_video(video_fp, audio_fp, wfp): + if osp.exists(video_fp) and osp.exists(audio_fp): + cmd = f'ffmpeg -i {video_fp} -i {audio_fp} -c:v copy -c:a aac {wfp} -y' + exec_cmd(cmd) + print(f'merge {video_fp} and {audio_fp} to {wfp}') + else: + print(f'video_fp: {video_fp} or audio_fp: {audio_fp} not exists!') + + +def blend(img: np.ndarray, mask: np.ndarray, background_color=(255, 255, 255)): + mask_float = mask.astype(np.float32) / 255. + background_color = np.array(background_color).reshape([1, 1, 3]) + bg = np.ones_like(img) * background_color + img = np.clip(mask_float * img + (1 - mask_float) * bg, 0, 255).astype(np.uint8) + return img + + +def concat_frames(I_p_lst, driving_rgb_lst, img_rgb): + # TODO: add more concat style, e.g., left-down corner driving + out_lst = [] + print('Concatenating result...',len(I_p_lst)) + for idx, _ in enumerate(I_p_lst): + # track(enumerate(I_p_lst), total=len(I_p_lst), description='Concatenating result...'): + source_image_drived = I_p_lst[idx] + image_drive = driving_rgb_lst[idx] + + # resize images to match source_image_drived shape + h, w, _ = source_image_drived.shape + image_drive_resized = cv2.resize(image_drive, (w, h)) + img_rgb_resized = cv2.resize(img_rgb, (w, h)) + + # concatenate images horizontally + frame = np.concatenate((image_drive_resized, img_rgb_resized, source_image_drived), axis=1) + out_lst.append(frame) + return out_lst + + +class VideoWriter: + def __init__(self, **kwargs): + self.fps = kwargs.get('fps', 30) + self.wfp = kwargs.get('wfp', 'video.mp4') + self.video_format = kwargs.get('format', 'mp4') + self.codec = kwargs.get('codec', 'libx264') + self.quality = kwargs.get('quality') + self.pixelformat = kwargs.get('pixelformat', 'yuv420p') + self.image_mode = kwargs.get('image_mode', 'rgb') + self.ffmpeg_params = kwargs.get('ffmpeg_params') + + self.writer = imageio.get_writer( + self.wfp, fps=self.fps, format=self.video_format, + codec=self.codec, quality=self.quality, + ffmpeg_params=self.ffmpeg_params, pixelformat=self.pixelformat + ) + + def write(self, image): + if self.image_mode.lower() == 'bgr': + self.writer.append_data(image[..., ::-1]) + else: + self.writer.append_data(image) + + def close(self): + if self.writer is not None: + self.writer.close() + + +def change_video_fps(input_file, output_file, fps=20, codec='libx264', crf=5): + cmd = f"ffmpeg -i {input_file} -c:v {codec} -crf {crf} -r {fps} {output_file} -y" + exec_cmd(cmd) + + +def get_fps(filepath): + import ffmpeg + probe = ffmpeg.probe(filepath) + video_stream = next((stream for stream in probe['streams'] if stream['codec_type'] == 'video'), None) + fps = eval(video_stream['avg_frame_rate']) + return fps diff --git a/ComfyUI-AdvancedLivePortrait/README.md b/ComfyUI-AdvancedLivePortrait/README.md new file mode 100644 index 0000000000000000000000000000000000000000..3b6ae54111a469d98b99ed184425931ed5997cc3 --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/README.md @@ -0,0 +1,64 @@ +# ComfyUI-AdvancedLivePortrait + +## Update + +8/21/2024 + +You can create a video without a video. + +Track the face of the source video. + +The workflow has been updated. + +## Introduction + +AdvancedLivePortrait is faster and has real-time preview + +https://github.com/user-attachments/assets/90b78639-6477-48af-ba49-7945488df581 + +Edit facial expressions in photos. + +Insert facial expressions into videos. + +Create animations using multiple facial expressions. + +Extract facial expressions from sample photos. + +## Installation + +This project has been registered with ComfyUI-Manager. Now you can install it automatically using the manager. + +## Usage + +The workflows and sample datas placed in '\custom_nodes\ComfyUI-AdvancedLivePortrait\sample\' + +----- + +You can add expressions to the video. See 'workflow2_advanced.json'. + +Describes the 'command' in 'workflow2_advanced.json' + +![readme](https://github.com/user-attachments/assets/339568b2-ad52-4aaf-a6ab-fcd877449c56) + + +[Motion index] = [Changing frame length] : [Length of frames waiting for next motion] + +Motion index 0 is the original source image. + +They are numbered in the order they lead to the motion_link. + +Linking the driving video to 'src_images' will add facial expressions to the driving video. + +----- + +You can save and load expressions with the 'Load Exp Data' 'Save Exp Data' nodes. + +\ComfyUI\output\exp_data\ Path to the folder being saved + +----- + +## Thanks + +Original author's link : https://liveportrait.github.io/ + +This project uses a model converted by kijai. link : https://github.com/kijai/ComfyUI-LivePortraitKJ diff --git a/ComfyUI-AdvancedLivePortrait/__init__.py b/ComfyUI-AdvancedLivePortrait/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..0e12fe01121421629b8b227881bfc78668883758 --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/__init__.py @@ -0,0 +1,4 @@ +from .nodes import NODE_CLASS_MAPPINGS, NODE_DISPLAY_NAME_MAPPINGS +__all__ = ["NODE_CLASS_MAPPINGS", "NODE_DISPLAY_NAME_MAPPINGS"] + + diff --git a/ComfyUI-AdvancedLivePortrait/__pycache__/__init__.cpython-312.pyc b/ComfyUI-AdvancedLivePortrait/__pycache__/__init__.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..1880e66ce3889a0bb34faaea28cb8bc6c25930a5 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/__pycache__/__init__.cpython-312.pyc differ diff --git a/ComfyUI-AdvancedLivePortrait/__pycache__/nodes.cpython-312.pyc b/ComfyUI-AdvancedLivePortrait/__pycache__/nodes.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..ac7ac504dbd5a447c5a29af5e06df232c631b73a Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/__pycache__/nodes.cpython-312.pyc differ diff --git a/ComfyUI-AdvancedLivePortrait/install.bat b/ComfyUI-AdvancedLivePortrait/install.bat new file mode 100644 index 0000000000000000000000000000000000000000..c26b782c27800dc17a579a090e88e9eb5a23e31f --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/install.bat @@ -0,0 +1,20 @@ +@echo off + +set "requirements_txt=%~dp0\requirements.txt" +set "python_exec=..\..\..\python_embeded\python.exe" + +echo Installing ComfyUI-AdvancedLivePortrait.. + +if exist "%python_exec%" ( + echo Installing with ComfyUI Portable + for /f "delims=" %%i in (%requirements_txt%) do ( + %python_exec% -s -m pip install "%%i" + ) +) else ( + echo Installing with system Python + for /f "delims=" %%i in (%requirements_txt%) do ( + pip install "%%i" + ) +) + +pause \ No newline at end of file diff --git a/ComfyUI-AdvancedLivePortrait/nodes.py b/ComfyUI-AdvancedLivePortrait/nodes.py new file mode 100644 index 0000000000000000000000000000000000000000..908892d28e18ac6a93f65a9066a1d2552b8ab42d --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/nodes.py @@ -0,0 +1,980 @@ +import os +import sys +import numpy as np +import torch +import cv2 +from PIL import Image +import folder_paths +import comfy.utils +import time +import copy +import dill +import yaml +from ultralytics import YOLO + +current_file_path = os.path.abspath(__file__) +current_directory = os.path.dirname(current_file_path) + +from .LivePortrait.live_portrait_wrapper import LivePortraitWrapper +from .LivePortrait.utils.camera import get_rotation_matrix +from .LivePortrait.config.inference_config import InferenceConfig + +from .LivePortrait.modules.spade_generator import SPADEDecoder +from .LivePortrait.modules.warping_network import WarpingNetwork +from .LivePortrait.modules.motion_extractor import MotionExtractor +from .LivePortrait.modules.appearance_feature_extractor import AppearanceFeatureExtractor +from .LivePortrait.modules.stitching_retargeting_network import StitchingRetargetingNetwork +from collections import OrderedDict + +cur_device = None +def get_device(): + global cur_device + if cur_device == None: + if torch.cuda.is_available(): + cur_device = torch.device('cuda') + print("Uses CUDA device.") + elif torch.backends.mps.is_available(): + cur_device = torch.device('mps') + print("Uses MPS device.") + else: + cur_device = torch.device('cpu') + print("Uses CPU device.") + return cur_device + +def tensor2pil(image): + return Image.fromarray(np.clip(255. * image.cpu().numpy().squeeze(), 0, 255).astype(np.uint8)) +def pil2tensor(image): + return torch.from_numpy(np.array(image).astype(np.float32) / 255.0).unsqueeze(0) +def rgb_crop(rgb, region): + return rgb[region[1]:region[3], region[0]:region[2]] + +def rgb_crop_batch(rgbs, region): + return rgbs[:, region[1]:region[3], region[0]:region[2]] +def get_rgb_size(rgb): + return rgb.shape[1], rgb.shape[0] +def create_transform_matrix(x, y, s_x, s_y): + return np.float32([[s_x, 0, x], [0, s_y, y]]) + +def get_model_dir(m): + try: + return folder_paths.get_folder_paths(m)[0] + except: + return os.path.join(folder_paths.models_dir, m) + +def calc_crop_limit(center, img_size, crop_size): + pos = center - crop_size / 2 + if pos < 0: + crop_size += pos * 2 + pos = 0 + + pos2 = pos + crop_size + + if img_size < pos2: + crop_size -= (pos2 - img_size) * 2 + pos2 = img_size + pos = pos2 - crop_size + + return pos, pos2, crop_size + +def retargeting(delta_out, driving_exp, factor, idxes): + for idx in idxes: + #delta_out[0, idx] -= src_exp[0, idx] * factor + delta_out[0, idx] += driving_exp[0, idx] * factor + +class PreparedSrcImg: + def __init__(self, src_rgb, crop_trans_m, x_s_info, f_s_user, x_s_user, mask_ori): + self.src_rgb = src_rgb + self.crop_trans_m = crop_trans_m + self.x_s_info = x_s_info + self.f_s_user = f_s_user + self.x_s_user = x_s_user + self.mask_ori = mask_ori + +import requests +from tqdm import tqdm + +class LP_Engine: + pipeline = None + detect_model = None + mask_img = None + temp_img_idx = 0 + + def get_temp_img_name(self): + self.temp_img_idx += 1 + return "expression_edit_preview" + str(self.temp_img_idx) + ".png" + + def download_model(_, file_path, model_url): + print('AdvancedLivePortrait: Downloading model...') + response = requests.get(model_url, stream=True) + try: + if response.status_code == 200: + total_size = int(response.headers.get('content-length', 0)) + block_size = 1024 # 1 Kibibyte + + # tqdm will display a progress bar + with open(file_path, 'wb') as file, tqdm( + desc='Downloading', + total=total_size, + unit='iB', + unit_scale=True, + unit_divisor=1024, + ) as bar: + for data in response.iter_content(block_size): + bar.update(len(data)) + file.write(data) + + except requests.exceptions.RequestException as err: + print('AdvancedLivePortrait: Model download failed: {err}') + print(f'AdvancedLivePortrait: Download it manually from: {model_url}') + print(f'AdvancedLivePortrait: And put it in {file_path}') + except Exception as e: + print(f'AdvancedLivePortrait: An unexpected error occurred: {e}') + + def remove_ddp_dumplicate_key(_, state_dict): + state_dict_new = OrderedDict() + for key in state_dict.keys(): + state_dict_new[key.replace('module.', '')] = state_dict[key] + return state_dict_new + + def filter_for_model(_, checkpoint, prefix): + filtered_checkpoint = {key.replace(prefix + "_module.", ""): value for key, value in checkpoint.items() if + key.startswith(prefix)} + return filtered_checkpoint + + def load_model(self, model_config, model_type): + + device = get_device() + + if model_type == 'stitching_retargeting_module': + ckpt_path = os.path.join(get_model_dir("liveportrait"), "retargeting_models", model_type + ".pth") + else: + ckpt_path = os.path.join(get_model_dir("liveportrait"), "base_models", model_type + ".pth") + + is_safetensors = None + if os.path.isfile(ckpt_path) == False: + is_safetensors = True + ckpt_path = os.path.join(get_model_dir("liveportrait"), model_type + ".safetensors") + if os.path.isfile(ckpt_path) == False: + self.download_model(ckpt_path, + "https://huggingface.co/Kijai/LivePortrait_safetensors/resolve/main/" + model_type + ".safetensors") + model_params = model_config['model_params'][f'{model_type}_params'] + if model_type == 'appearance_feature_extractor': + model = AppearanceFeatureExtractor(**model_params).to(device) + elif model_type == 'motion_extractor': + model = MotionExtractor(**model_params).to(device) + elif model_type == 'warping_module': + model = WarpingNetwork(**model_params).to(device) + elif model_type == 'spade_generator': + model = SPADEDecoder(**model_params).to(device) + elif model_type == 'stitching_retargeting_module': + # Special handling for stitching and retargeting module + config = model_config['model_params']['stitching_retargeting_module_params'] + checkpoint = comfy.utils.load_torch_file(ckpt_path) + + stitcher = StitchingRetargetingNetwork(**config.get('stitching')) + if is_safetensors: + stitcher.load_state_dict(self.filter_for_model(checkpoint, 'retarget_shoulder')) + else: + stitcher.load_state_dict(self.remove_ddp_dumplicate_key(checkpoint['retarget_shoulder'])) + stitcher = stitcher.to(device) + stitcher.eval() + + return { + 'stitching': stitcher, + } + else: + raise ValueError(f"Unknown model type: {model_type}") + + + model.load_state_dict(comfy.utils.load_torch_file(ckpt_path)) + model.eval() + return model + + def load_models(self): + model_path = get_model_dir("liveportrait") + if not os.path.exists(model_path): + os.mkdir(model_path) + + model_config_path = os.path.join(current_directory, 'LivePortrait', 'config', 'models.yaml') + model_config = yaml.safe_load(open(model_config_path, 'r')) + + appearance_feature_extractor = self.load_model(model_config, 'appearance_feature_extractor') + motion_extractor = self.load_model(model_config, 'motion_extractor') + warping_module = self.load_model(model_config, 'warping_module') + spade_generator = self.load_model(model_config, 'spade_generator') + stitching_retargeting_module = self.load_model(model_config, 'stitching_retargeting_module') + + self.pipeline = LivePortraitWrapper(InferenceConfig(), appearance_feature_extractor, motion_extractor, warping_module, spade_generator, stitching_retargeting_module) + + def get_detect_model(self): + if self.detect_model == None: + model_dir = get_model_dir("ultralytics") + if not os.path.exists(model_dir): os.mkdir(model_dir) + model_path = os.path.join(model_dir, "face_yolov8n.pt") + if not os.path.exists(model_path): + self.download_model(model_path, "https://huggingface.co/Bingsu/adetailer/resolve/main/face_yolov8n.pt") + self.detect_model = YOLO(model_path) + + return self.detect_model + + def get_face_bboxes(self, image_rgb): + detect_model = self.get_detect_model() + pred = detect_model(image_rgb, conf=0.7, device="") + return pred[0].boxes.xyxy.cpu().numpy() + + def detect_face(self, image_rgb, crop_factor, sort = True): + bboxes = self.get_face_bboxes(image_rgb) + w, h = get_rgb_size(image_rgb) + + print(f"w, h:{w, h}") + + cx = w / 2 + min_diff = w + best_box = None + for x1, y1, x2, y2 in bboxes: + bbox_w = x2 - x1 + if bbox_w < 30: continue + diff = abs(cx - (x1 + bbox_w / 2)) + if diff < min_diff: + best_box = [x1, y1, x2, y2] + print(f"diff, min_diff, best_box:{diff, min_diff, best_box}") + min_diff = diff + + if best_box == None: + print("Failed to detect face!!") + return [0, 0, w, h] + + x1, y1, x2, y2 = best_box + + #for x1, y1, x2, y2 in bboxes: + bbox_w = x2 - x1 + bbox_h = y2 - y1 + + crop_w = bbox_w * crop_factor + crop_h = bbox_h * crop_factor + + crop_w = max(crop_h, crop_w) + crop_h = crop_w + + kernel_x = int(x1 + bbox_w / 2) + kernel_y = int(y1 + bbox_h / 2) + + new_x1 = int(kernel_x - crop_w / 2) + new_x2 = int(kernel_x + crop_w / 2) + new_y1 = int(kernel_y - crop_h / 2) + new_y2 = int(kernel_y + crop_h / 2) + + if not sort: + return [int(new_x1), int(new_y1), int(new_x2), int(new_y2)] + + if new_x1 < 0: + new_x2 -= new_x1 + new_x1 = 0 + elif w < new_x2: + new_x1 -= (new_x2 - w) + new_x2 = w + if new_x1 < 0: + new_x2 -= new_x1 + new_x1 = 0 + + if new_y1 < 0: + new_y2 -= new_y1 + new_y1 = 0 + elif h < new_y2: + new_y1 -= (new_y2 - h) + new_y2 = h + if new_y1 < 0: + new_y2 -= new_y1 + new_y1 = 0 + + if w < new_x2 and h < new_y2: + over_x = new_x2 - w + over_y = new_y2 - h + over_min = min(over_x, over_y) + new_x2 -= over_min + new_y2 -= over_min + + return [int(new_x1), int(new_y1), int(new_x2), int(new_y2)] + + + def calc_face_region(self, square, dsize): + region = copy.deepcopy(square) + is_changed = False + if dsize[0] < region[2]: + region[2] = dsize[0] + is_changed = True + if dsize[1] < region[3]: + region[3] = dsize[1] + is_changed = True + + return region, is_changed + + def expand_img(self, rgb_img, square): + #new_img = rgb_crop(rgb_img, face_region) + crop_trans_m = create_transform_matrix(max(-square[0], 0), max(-square[1], 0), 1, 1) + new_img = cv2.warpAffine(rgb_img, crop_trans_m, (square[2] - square[0], square[3] - square[1]), + cv2.INTER_LINEAR) + return new_img + + def get_pipeline(self): + if self.pipeline == None: + print("Load pipeline...") + self.load_models() + + return self.pipeline + + def prepare_src_image(self, img): + h, w = img.shape[:2] + input_shape = [256,256] + if h != input_shape[0] or w != input_shape[1]: + if 256 < h: interpolation = cv2.INTER_AREA + else: interpolation = cv2.INTER_LINEAR + x = cv2.resize(img, (input_shape[0], input_shape[1]), interpolation = interpolation) + else: + x = img.copy() + + if x.ndim == 3: + x = x[np.newaxis].astype(np.float32) / 255. # HxWx3 -> 1xHxWx3, normalized to 0~1 + elif x.ndim == 4: + x = x.astype(np.float32) / 255. # BxHxWx3, normalized to 0~1 + else: + raise ValueError(f'img ndim should be 3 or 4: {x.ndim}') + x = np.clip(x, 0, 1) # clip to 0~1 + x = torch.from_numpy(x).permute(0, 3, 1, 2) # 1xHxWx3 -> 1x3xHxW + x = x.to(get_device()) + return x + + def GetMaskImg(self): + if self.mask_img is None: + path = os.path.join(os.path.dirname(os.path.realpath(__file__)), "./LivePortrait/utils/resources/mask_template.png") + self.mask_img = cv2.imread(path, cv2.IMREAD_COLOR) + return self.mask_img + + def crop_face(self, img_rgb, crop_factor): + crop_region = self.detect_face(img_rgb, crop_factor) + face_region, is_changed = self.calc_face_region(crop_region, get_rgb_size(img_rgb)) + face_img = rgb_crop(img_rgb, face_region) + if is_changed: face_img = self.expand_img(face_img, crop_region) + return face_img + + def prepare_source(self, source_image, crop_factor, is_video = False, tracking = False): + print("Prepare source...") + engine = self.get_pipeline() + source_image_np = (source_image * 255).byte().numpy() + img_rgb = source_image_np[0] + + psi_list = [] + for img_rgb in source_image_np: + if tracking or len(psi_list) == 0: + crop_region = self.detect_face(img_rgb, crop_factor) + face_region, is_changed = self.calc_face_region(crop_region, get_rgb_size(img_rgb)) + + s_x = (face_region[2] - face_region[0]) / 512. + s_y = (face_region[3] - face_region[1]) / 512. + crop_trans_m = create_transform_matrix(crop_region[0], crop_region[1], s_x, s_y) + mask_ori = cv2.warpAffine(self.GetMaskImg(), crop_trans_m, get_rgb_size(img_rgb), cv2.INTER_LINEAR) + mask_ori = mask_ori.astype(np.float32) / 255. + + if is_changed: + s = (crop_region[2] - crop_region[0]) / 512. + crop_trans_m = create_transform_matrix(crop_region[0], crop_region[1], s, s) + + face_img = rgb_crop(img_rgb, face_region) + if is_changed: face_img = self.expand_img(face_img, crop_region) + i_s = self.prepare_src_image(face_img) + x_s_info = engine.get_kp_info(i_s) + f_s_user = engine.extract_feature_3d(i_s) + x_s_user = engine.transform_keypoint(x_s_info) + psi = PreparedSrcImg(img_rgb, crop_trans_m, x_s_info, f_s_user, x_s_user, mask_ori) + if is_video == False: + return psi + psi_list.append(psi) + + return psi_list + + def prepare_driving_video(self, face_images): + print("Prepare driving video...") + pipeline = self.get_pipeline() + f_img_np = (face_images * 255).byte().numpy() + + out_list = [] + for f_img in f_img_np: + i_d = self.prepare_src_image(f_img) + d_info = pipeline.get_kp_info(i_d) + out_list.append(d_info) + + return out_list + + def calc_fe(_, x_d_new, eyes, eyebrow, wink, pupil_x, pupil_y, mouth, eee, woo, smile, + rotate_pitch, rotate_yaw, rotate_roll): + + x_d_new[0, 20, 1] += smile * -0.01 + x_d_new[0, 14, 1] += smile * -0.02 + x_d_new[0, 17, 1] += smile * 0.0065 + x_d_new[0, 17, 2] += smile * 0.003 + x_d_new[0, 13, 1] += smile * -0.00275 + x_d_new[0, 16, 1] += smile * -0.00275 + x_d_new[0, 3, 1] += smile * -0.0035 + x_d_new[0, 7, 1] += smile * -0.0035 + + x_d_new[0, 19, 1] += mouth * 0.001 + x_d_new[0, 19, 2] += mouth * 0.0001 + x_d_new[0, 17, 1] += mouth * -0.0001 + rotate_pitch -= mouth * 0.05 + + x_d_new[0, 20, 2] += eee * -0.001 + x_d_new[0, 20, 1] += eee * -0.001 + #x_d_new[0, 19, 1] += eee * 0.0006 + x_d_new[0, 14, 1] += eee * -0.001 + + x_d_new[0, 14, 1] += woo * 0.001 + x_d_new[0, 3, 1] += woo * -0.0005 + x_d_new[0, 7, 1] += woo * -0.0005 + x_d_new[0, 17, 2] += woo * -0.0005 + + x_d_new[0, 11, 1] += wink * 0.001 + x_d_new[0, 13, 1] += wink * -0.0003 + x_d_new[0, 17, 0] += wink * 0.0003 + x_d_new[0, 17, 1] += wink * 0.0003 + x_d_new[0, 3, 1] += wink * -0.0003 + rotate_roll -= wink * 0.1 + rotate_yaw -= wink * 0.1 + + if 0 < pupil_x: + x_d_new[0, 11, 0] += pupil_x * 0.0007 + x_d_new[0, 15, 0] += pupil_x * 0.001 + else: + x_d_new[0, 11, 0] += pupil_x * 0.001 + x_d_new[0, 15, 0] += pupil_x * 0.0007 + + x_d_new[0, 11, 1] += pupil_y * -0.001 + x_d_new[0, 15, 1] += pupil_y * -0.001 + eyes -= pupil_y / 2. + + x_d_new[0, 11, 1] += eyes * -0.001 + x_d_new[0, 13, 1] += eyes * 0.0003 + x_d_new[0, 15, 1] += eyes * -0.001 + x_d_new[0, 16, 1] += eyes * 0.0003 + x_d_new[0, 1, 1] += eyes * -0.00025 + x_d_new[0, 2, 1] += eyes * 0.00025 + + + if 0 < eyebrow: + x_d_new[0, 1, 1] += eyebrow * 0.001 + x_d_new[0, 2, 1] += eyebrow * -0.001 + else: + x_d_new[0, 1, 0] += eyebrow * -0.001 + x_d_new[0, 2, 0] += eyebrow * 0.001 + x_d_new[0, 1, 1] += eyebrow * 0.0003 + x_d_new[0, 2, 1] += eyebrow * -0.0003 + + + return torch.Tensor([rotate_pitch, rotate_yaw, rotate_roll]) +g_engine = LP_Engine() + +class ExpressionSet: + def __init__(self, erst = None, es = None): + if es != None: + self.e = copy.deepcopy(es.e) # [:, :, :] + self.r = copy.deepcopy(es.r) # [:] + self.s = copy.deepcopy(es.s) + self.t = copy.deepcopy(es.t) + elif erst != None: + self.e = erst[0] + self.r = erst[1] + self.s = erst[2] + self.t = erst[3] + else: + self.e = torch.from_numpy(np.zeros((1, 21, 3))).float().to(get_device()) + self.r = torch.Tensor([0, 0, 0]) + self.s = 0 + self.t = 0 + def div(self, value): + self.e /= value + self.r /= value + self.s /= value + self.t /= value + def add(self, other): + self.e += other.e + self.r += other.r + self.s += other.s + self.t += other.t + def sub(self, other): + self.e -= other.e + self.r -= other.r + self.s -= other.s + self.t -= other.t + def mul(self, value): + self.e *= value + self.r *= value + self.s *= value + self.t *= value + + #def apply_ratio(self, ratio): self.exp *= ratio + +def logging_time(original_fn): + def wrapper_fn(*args, **kwargs): + start_time = time.time() + result = original_fn(*args, **kwargs) + end_time = time.time() + print("WorkingTime[{}]: {} sec".format(original_fn.__name__, end_time - start_time)) + return result + + return wrapper_fn + + +#exp_data_dir = os.path.join(current_directory, "exp_data") +exp_data_dir = os.path.join(folder_paths.output_directory, "exp_data") +if os.path.isdir(exp_data_dir) == False: + os.mkdir(exp_data_dir) +class SaveExpData: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "file_name": ("STRING", {"multiline": False, "default": ""}), + }, + "optional": {"save_exp": ("EXP_DATA",), } + } + + RETURN_TYPES = ("STRING",) + RETURN_NAMES = ("file_name",) + FUNCTION = "run" + CATEGORY = "AdvancedLivePortrait" + OUTPUT_NODE = True + + def run(self, file_name, save_exp:ExpressionSet=None): + if save_exp == None or file_name == "": + return file_name + + with open(os.path.join(exp_data_dir, file_name + ".exp"), "wb") as f: + dill.dump(save_exp, f) + + return file_name + +class LoadExpData: + @classmethod + def INPUT_TYPES(s): + file_list = [os.path.splitext(file)[0] for file in os.listdir(exp_data_dir) if file.endswith('.exp')] + return {"required": { + "file_name": (sorted(file_list, key=str.lower),), + "ratio": ("FLOAT", {"default": 1, "min": 0, "max": 1, "step": 0.01}), + }, + } + + RETURN_TYPES = ("EXP_DATA",) + RETURN_NAMES = ("exp",) + FUNCTION = "run" + CATEGORY = "AdvancedLivePortrait" + + def run(self, file_name, ratio): + # es = ExpressionSet() + with open(os.path.join(exp_data_dir, file_name + ".exp"), 'rb') as f: + es = dill.load(f) + es.mul(ratio) + return (es,) + +class ExpData: + @classmethod + def INPUT_TYPES(s): + return {"required":{ + #"code": ("STRING", {"multiline": False, "default": ""}), + "code1": ("INT", {"default": 0}), + "value1": ("FLOAT", {"default": 0, "min": -100, "max": 100, "step": 0.1}), + "code2": ("INT", {"default": 0}), + "value2": ("FLOAT", {"default": 0, "min": -100, "max": 100, "step": 0.1}), + "code3": ("INT", {"default": 0}), + "value3": ("FLOAT", {"default": 0, "min": -100, "max": 100, "step": 0.1}), + "code4": ("INT", {"default": 0}), + "value4": ("FLOAT", {"default": 0, "min": -100, "max": 100, "step": 0.1}), + "code5": ("INT", {"default": 0}), + "value5": ("FLOAT", {"default": 0, "min": -100, "max": 100, "step": 0.1}), + }, + "optional":{"add_exp": ("EXP_DATA",),} + } + + RETURN_TYPES = ("EXP_DATA",) + RETURN_NAMES = ("exp",) + FUNCTION = "run" + CATEGORY = "AdvancedLivePortrait" + + def run(self, code1, value1, code2, value2, code3, value3, code4, value4, code5, value5, add_exp=None): + if add_exp == None: + es = ExpressionSet() + else: + es = ExpressionSet(es = add_exp) + + codes = [code1, code2, code3, code4, code5] + values = [value1, value2, value3, value4, value5] + for i in range(5): + idx = int(codes[i] / 10) + r = codes[i] % 10 + es.e[0, idx, r] += values[i] * 0.001 + + return (es,) + +class PrintExpData: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "cut_noise": ("FLOAT", {"default": 0, "min": 0, "max": 100, "step": 0.1}), + }, + "optional": {"exp": ("EXP_DATA",), } + } + + RETURN_TYPES = ("EXP_DATA",) + RETURN_NAMES = ("exp",) + FUNCTION = "run" + CATEGORY = "AdvancedLivePortrait" + OUTPUT_NODE = True + + def run(self, cut_noise, exp = None): + if exp == None: return (exp,) + + cuted_list = [] + e = exp.exp * 1000 + for idx in range(21): + for r in range(3): + a = abs(e[0, idx, r]) + if(cut_noise < a): cuted_list.append((a, e[0, idx, r], idx*10+r)) + + sorted_list = sorted(cuted_list, reverse=True, key=lambda item: item[0]) + print(f"sorted_list: {[[item[2], round(float(item[1]),1)] for item in sorted_list]}") + return (exp,) + +class Command: + def __init__(self, es, change, keep): + self.es:ExpressionSet = es + self.change = change + self.keep = keep + +crop_factor_default = 1.7 +crop_factor_min = 1.5 +crop_factor_max = 2.5 + +class AdvancedLivePortrait: + def __init__(self): + self.src_images = None + self.driving_images = None + self.pbar = comfy.utils.ProgressBar(1) + self.crop_factor = None + + @classmethod + def INPUT_TYPES(s): + + return { + "required": { + "retargeting_eyes": ("FLOAT", {"default": 0, "min": 0, "max": 1, "step": 0.01}), + "retargeting_mouth": ("FLOAT", {"default": 0, "min": 0, "max": 1, "step": 0.01}), + "crop_factor": ("FLOAT", {"default": crop_factor_default, + "min": crop_factor_min, "max": crop_factor_max, "step": 0.1}), + "turn_on": ("BOOLEAN", {"default": True}), + "tracking_src_vid": ("BOOLEAN", {"default": False}), + "animate_without_vid": ("BOOLEAN", {"default": False}), + "command": ("STRING", {"multiline": True, "default": ""}), + }, + "optional": { + "src_images": ("IMAGE",), + "motion_link": ("EDITOR_LINK",), + "driving_images": ("IMAGE",), + }, + } + + RETURN_TYPES = ("IMAGE",) + RETURN_NAMES = ("images",) + FUNCTION = "run" + OUTPUT_NODE = True + CATEGORY = "AdvancedLivePortrait" + + # INPUT_IS_LIST = False + # OUTPUT_IS_LIST = (False,) + + def parsing_command(self, command, motoin_link): + command.replace(' ', '') + # if command == '': return + lines = command.split('\n') + + cmd_list = [] + + total_length = 0 + + i = 0 + #old_es = None + for line in lines: + i += 1 + if line == '': continue + try: + cmds = line.split('=') + idx = int(cmds[0]) + if idx == 0: es = ExpressionSet() + else: es = ExpressionSet(es = motoin_link[idx]) + cmds = cmds[1].split(':') + change = int(cmds[0]) + keep = int(cmds[1]) + except: + assert False, f"(AdvancedLivePortrait) Command Err Line {i}: {line}" + + + return None, None + + total_length += change + keep + es.div(change) + cmd_list.append(Command(es, change, keep)) + + return cmd_list, total_length + + + def run(self, retargeting_eyes, retargeting_mouth, turn_on, tracking_src_vid, animate_without_vid, command, crop_factor, + src_images=None, driving_images=None, motion_link=None): + if turn_on == False: return (None,None) + src_length = 1 + + if src_images == None: + if motion_link != None: + self.psi_list = [motion_link[0]] + else: return (None,None) + + if src_images != None: + src_length = len(src_images) + if id(src_images) != id(self.src_images) or self.crop_factor != crop_factor: + self.crop_factor = crop_factor + self.src_images = src_images + if 1 < src_length: + self.psi_list = g_engine.prepare_source(src_images, crop_factor, True, tracking_src_vid) + else: + self.psi_list = [g_engine.prepare_source(src_images, crop_factor)] + + + cmd_list, cmd_length = self.parsing_command(command, motion_link) + if cmd_list == None: return (None,None) + cmd_idx = 0 + + driving_length = 0 + if driving_images is not None: + if id(driving_images) != id(self.driving_images): + self.driving_images = driving_images + self.driving_values = g_engine.prepare_driving_video(driving_images) + driving_length = len(self.driving_values) + + total_length = max(driving_length, src_length) + + if animate_without_vid: + total_length = max(total_length, cmd_length) + + c_i_es = ExpressionSet() + c_o_es = ExpressionSet() + d_0_es = None + out_list = [] + + psi = None + pipeline = g_engine.get_pipeline() + for i in range(total_length): + + if i < src_length: + psi = self.psi_list[i] + s_info = psi.x_s_info + s_es = ExpressionSet(erst=(s_info['kp'] + s_info['exp'], torch.Tensor([0, 0, 0]), s_info['scale'], s_info['t'])) + + new_es = ExpressionSet(es = s_es) + + if i < cmd_length: + cmd = cmd_list[cmd_idx] + if 0 < cmd.change: + cmd.change -= 1 + c_i_es.add(cmd.es) + c_i_es.sub(c_o_es) + elif 0 < cmd.keep: + cmd.keep -= 1 + + new_es.add(c_i_es) + + if cmd.change == 0 and cmd.keep == 0: + cmd_idx += 1 + if cmd_idx < len(cmd_list): + c_o_es = ExpressionSet(es = c_i_es) + cmd = cmd_list[cmd_idx] + c_o_es.div(cmd.change) + elif 0 < cmd_length: + new_es.add(c_i_es) + + if i < driving_length: + d_i_info = self.driving_values[i] + d_i_r = torch.Tensor([d_i_info['pitch'], d_i_info['yaw'], d_i_info['roll']])#.float().to(device="cuda:0") + + if d_0_es is None: + d_0_es = ExpressionSet(erst = (d_i_info['exp'], d_i_r, d_i_info['scale'], d_i_info['t'])) + + retargeting(s_es.e, d_0_es.e, retargeting_eyes, (11, 13, 15, 16)) + retargeting(s_es.e, d_0_es.e, retargeting_mouth, (14, 17, 19, 20)) + + new_es.e += d_i_info['exp'] - d_0_es.e + new_es.r += d_i_r - d_0_es.r + new_es.t += d_i_info['t'] - d_0_es.t + + r_new = get_rotation_matrix( + s_info['pitch'] + new_es.r[0], s_info['yaw'] + new_es.r[1], s_info['roll'] + new_es.r[2]) + d_new = new_es.s * (new_es.e @ r_new) + new_es.t + d_new = pipeline.stitching(psi.x_s_user, d_new) + crop_out = pipeline.warp_decode(psi.f_s_user, psi.x_s_user, d_new) + crop_out = pipeline.parse_output(crop_out['out'])[0] + + crop_with_fullsize = cv2.warpAffine(crop_out, psi.crop_trans_m, get_rgb_size(psi.src_rgb), + cv2.INTER_LINEAR) + out = np.clip(psi.mask_ori * crop_with_fullsize + (1 - psi.mask_ori) * psi.src_rgb, 0, 255).astype( + np.uint8) + out_list.append(out) + + self.pbar.update_absolute(i+1, total_length, ("PNG", Image.fromarray(crop_out), None)) + + if len(out_list) == 0: return (None,) + + out_imgs = torch.cat([pil2tensor(img_rgb) for img_rgb in out_list]) + return (out_imgs,) + +class ExpressionEditor: + def __init__(self): + self.sample_image = None + self.src_image = None + self.crop_factor = None + + @classmethod + def INPUT_TYPES(s): + display = "number" + #display = "slider" + return { + "required": { + + "rotate_pitch": ("FLOAT", {"default": 0, "min": -20, "max": 20, "step": 0.5, "display": display}), + "rotate_yaw": ("FLOAT", {"default": 0, "min": -20, "max": 20, "step": 0.5, "display": display}), + "rotate_roll": ("FLOAT", {"default": 0, "min": -20, "max": 20, "step": 0.5, "display": display}), + + "blink": ("FLOAT", {"default": 0, "min": -20, "max": 5, "step": 0.5, "display": display}), + "eyebrow": ("FLOAT", {"default": 0, "min": -10, "max": 15, "step": 0.5, "display": display}), + "wink": ("FLOAT", {"default": 0, "min": 0, "max": 25, "step": 0.5, "display": display}), + "pupil_x": ("FLOAT", {"default": 0, "min": -15, "max": 15, "step": 0.5, "display": display}), + "pupil_y": ("FLOAT", {"default": 0, "min": -15, "max": 15, "step": 0.5, "display": display}), + "aaa": ("FLOAT", {"default": 0, "min": -30, "max": 120, "step": 1, "display": display}), + "eee": ("FLOAT", {"default": 0, "min": -20, "max": 15, "step": 0.2, "display": display}), + "woo": ("FLOAT", {"default": 0, "min": -20, "max": 15, "step": 0.2, "display": display}), + "smile": ("FLOAT", {"default": 0, "min": -0.3, "max": 1.3, "step": 0.01, "display": display}), + + "src_ratio": ("FLOAT", {"default": 1, "min": 0, "max": 1, "step": 0.01, "display": display}), + "sample_ratio": ("FLOAT", {"default": 1, "min": -0.2, "max": 1.2, "step": 0.01, "display": display}), + "sample_parts": (["OnlyExpression", "OnlyRotation", "OnlyMouth", "OnlyEyes", "All"],), + "crop_factor": ("FLOAT", {"default": crop_factor_default, + "min": crop_factor_min, "max": crop_factor_max, "step": 0.1}), + }, + + "optional": {"src_image": ("IMAGE",), "motion_link": ("EDITOR_LINK",), + "sample_image": ("IMAGE",), "add_exp": ("EXP_DATA",), + }, + } + + RETURN_TYPES = ("IMAGE", "EDITOR_LINK", "EXP_DATA") + RETURN_NAMES = ("image", "motion_link", "save_exp") + + FUNCTION = "run" + + OUTPUT_NODE = True + + CATEGORY = "AdvancedLivePortrait" + + # INPUT_IS_LIST = False + # OUTPUT_IS_LIST = (False,) + + def run(self, rotate_pitch, rotate_yaw, rotate_roll, blink, eyebrow, wink, pupil_x, pupil_y, aaa, eee, woo, smile, + src_ratio, sample_ratio, sample_parts, crop_factor, src_image=None, sample_image=None, motion_link=None, add_exp=None): + rotate_yaw = -rotate_yaw + + new_editor_link = None + if motion_link != None: + self.psi = motion_link[0] + new_editor_link = motion_link.copy() + elif src_image != None: + if id(src_image) != id(self.src_image) or self.crop_factor != crop_factor: + self.crop_factor = crop_factor + self.psi = g_engine.prepare_source(src_image, crop_factor) + self.src_image = src_image + new_editor_link = [] + new_editor_link.append(self.psi) + else: + return (None,None) + + pipeline = g_engine.get_pipeline() + + psi = self.psi + s_info = psi.x_s_info + #delta_new = copy.deepcopy() + s_exp = s_info['exp'] * src_ratio + s_exp[0, 5] = s_info['exp'][0, 5] + s_exp += s_info['kp'] + + es = ExpressionSet() + + if sample_image != None: + if id(self.sample_image) != id(sample_image): + self.sample_image = sample_image + d_image_np = (sample_image * 255).byte().numpy() + d_face = g_engine.crop_face(d_image_np[0], 1.7) + i_d = g_engine.prepare_src_image(d_face) + self.d_info = pipeline.get_kp_info(i_d) + self.d_info['exp'][0, 5, 0] = 0 + self.d_info['exp'][0, 5, 1] = 0 + + # "OnlyExpression", "OnlyRotation", "OnlyMouth", "OnlyEyes", "All" + if sample_parts == "OnlyExpression" or sample_parts == "All": + es.e += self.d_info['exp'] * sample_ratio + if sample_parts == "OnlyRotation" or sample_parts == "All": + rotate_pitch += self.d_info['pitch'] * sample_ratio + rotate_yaw += self.d_info['yaw'] * sample_ratio + rotate_roll += self.d_info['roll'] * sample_ratio + elif sample_parts == "OnlyMouth": + retargeting(es.e, self.d_info['exp'], sample_ratio, (14, 17, 19, 20)) + elif sample_parts == "OnlyEyes": + retargeting(es.e, self.d_info['exp'], sample_ratio, (1, 2, 11, 13, 15, 16)) + + es.r = g_engine.calc_fe(es.e, blink, eyebrow, wink, pupil_x, pupil_y, aaa, eee, woo, smile, + rotate_pitch, rotate_yaw, rotate_roll) + + if add_exp != None: + es.add(add_exp) + + new_rotate = get_rotation_matrix(s_info['pitch'] + es.r[0], s_info['yaw'] + es.r[1], + s_info['roll'] + es.r[2]) + x_d_new = (s_info['scale'] * (1 + es.s)) * ((s_exp + es.e) @ new_rotate) + s_info['t'] + + x_d_new = pipeline.stitching(psi.x_s_user, x_d_new) + + crop_out = pipeline.warp_decode(psi.f_s_user, psi.x_s_user, x_d_new) + crop_out = pipeline.parse_output(crop_out['out'])[0] + + crop_with_fullsize = cv2.warpAffine(crop_out, psi.crop_trans_m, get_rgb_size(psi.src_rgb), cv2.INTER_LINEAR) + out = np.clip(psi.mask_ori * crop_with_fullsize + (1 - psi.mask_ori) * psi.src_rgb, 0, 255).astype(np.uint8) + + out_img = pil2tensor(out) + + filename = g_engine.get_temp_img_name() #"fe_edit_preview.png" + folder_paths.get_save_image_path(filename, folder_paths.get_temp_directory()) + img = Image.fromarray(crop_out) + img.save(os.path.join(folder_paths.get_temp_directory(), filename), compress_level=1) + results = list() + results.append({"filename": filename, "type": "temp"}) + + new_editor_link.append(es) + + return {"ui": {"images": results}, "result": (out_img, new_editor_link, es)} + +NODE_CLASS_MAPPINGS = { + "AdvancedLivePortrait": AdvancedLivePortrait, + "ExpressionEditor": ExpressionEditor, + "LoadExpData": LoadExpData, + "SaveExpData": SaveExpData, + "ExpData": ExpData, + "PrintExpData:": PrintExpData, +} + +NODE_DISPLAY_NAME_MAPPINGS = { + "AdvancedLivePortrait": "Advanced Live Portrait (PHM)", + "ExpressionEditor": "Expression Editor (PHM)", + "LoadExpData": "Load Exp Data (PHM)", + "SaveExpData": "Save Exp Data (PHM)" +} \ No newline at end of file diff --git a/ComfyUI-AdvancedLivePortrait/pyproject.toml b/ComfyUI-AdvancedLivePortrait/pyproject.toml new file mode 100644 index 0000000000000000000000000000000000000000..5db4f3f3d35aac782c2e7a15ebcee63a371bb1d1 --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/pyproject.toml @@ -0,0 +1,15 @@ +[project] +name = "comfyui-advancedliveportrait" +description = "AdvancedLivePortrait with Facial expression editor" +version = "1.0.0" +license = {file = "LICENSE"} +dependencies = ["numpy>=1.26.4", "opencv-python-headless", "imageio-ffmpeg>=0.5.1", "lmdb>=1.4.1", "timm>=1.0.7", "rich>=13.7.1", "albumentations>=1.4.10", "ultralytics", "tyro==0.8.5", "dill"] + +[project.urls] +Repository = "https://github.com/PowerHouseMan/ComfyUI-AdvancedLivePortrait" +# Used by Comfy Registry https://comfyregistry.org + +[tool.comfy] +PublisherId = "starmapking" +DisplayName = "ComfyUI-AdvancedLivePortrait" +Icon = "" diff --git a/ComfyUI-AdvancedLivePortrait/requirements.txt b/ComfyUI-AdvancedLivePortrait/requirements.txt new file mode 100644 index 0000000000000000000000000000000000000000..f4f7aa147831266fd6c06f8ec2f129d3e34512dd --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/requirements.txt @@ -0,0 +1,10 @@ +numpy>=1.26.4 +opencv-python-headless +imageio-ffmpeg>=0.5.1 +lmdb>=1.4.1 +timm>=1.0.7 +rich>=13.7.1 +albumentations>=1.4.10 +ultralytics +tyro==0.8.5 +dill \ No newline at end of file diff --git a/ComfyUI-AdvancedLivePortrait/sample/driving_video.mp4 b/ComfyUI-AdvancedLivePortrait/sample/driving_video.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..81b5ae1a6f301d613100547cce5ba06eeb033e12 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/driving_video.mp4 differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/exp_image.png b/ComfyUI-AdvancedLivePortrait/sample/exp_image.png new file mode 100644 index 0000000000000000000000000000000000000000..b0a719bc9854f5208aafa008aef0ddc34883bf59 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/exp_image.png differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d0.mp4 b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d0.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..92391dd3ff235fc82f29b7cc77fe4a7ce183d934 --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d0.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:63f6f9962e1fdf6e6722172e7a18155204858d5d5ce3b1e0646c150360c33bed +size 2958395 diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d1.mp4 b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d1.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..e2825c1d0663cb82aadc035f2cfbc9c75288a766 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d1.mp4 differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d2.mp4 b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d2.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..a14da2d8a07d6b88f4581635052982c15b15d988 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d2.mp4 differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d3.mp4 b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d3.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..8b70b6aa3c0e566a4fa3e5959f2d3b916e99b708 --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d3.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ef5c86e49b1b43dcb1449b499eb5a7f0cbae2f78aec08b5598193be1e4257099 +size 1430968 diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d5.mp4 b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d5.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..332bc88cc665df3135662086f1b916b3789b380a Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d5.mp4 differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d6.mp4 b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d6.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..44f351385cef843b21b03fab8c3b10e0c005ec5e --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d6.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:00e3ea79bbf28cbdc4fbb67ec655d9a0fe876e880ec45af55ae481348d0c0fff +size 1967790 diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d7.mp4 b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d7.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..81b5ae1a6f301d613100547cce5ba06eeb033e12 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d7.mp4 differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d8.mp4 b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d8.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..7fabddedb916f57eb35d114b0a9c88c7a212d68f Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d8.mp4 differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d9.mp4 b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d9.mp4 new file mode 100644 index 0000000000000000000000000000000000000000..7803b3bf5c460a79d94e5cfbedb0de1f52d449d2 --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/driving/d9.mp4 @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9a414aa1d547be35306d692065a2157434bf40a6025ba8e30ce12e5bb322cc33 +size 2257929 diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s0.jpg b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s0.jpg new file mode 100644 index 0000000000000000000000000000000000000000..ef44c593be38cea30422fff9ed986a8a77889348 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s0.jpg differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s1.jpg b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s1.jpg new file mode 100644 index 0000000000000000000000000000000000000000..ebacda3519a1452aee239f7e104d2c6ff40beb25 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s1.jpg differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s10.jpg b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s10.jpg new file mode 100644 index 0000000000000000000000000000000000000000..ee9616b592f070fbe90a8717da01477e8d4ee01f Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s10.jpg differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s2.jpg b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s2.jpg new file mode 100644 index 0000000000000000000000000000000000000000..e851bd20b65c552266a87bb87a9b509e3ea56f7d Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s2.jpg differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s3.jpg b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s3.jpg new file mode 100644 index 0000000000000000000000000000000000000000..9f3ba2a358e5b88450e7466761dff3e983e18e16 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s3.jpg differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s4.jpg b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s4.jpg new file mode 100644 index 0000000000000000000000000000000000000000..17f611bf942ad168d4e4d03b7e5c42d6650c4be1 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s4.jpg differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s5.jpg b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s5.jpg new file mode 100644 index 0000000000000000000000000000000000000000..9abad7ef061b93579a373cf141d38710d9b1e32d Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s5.jpg differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s6.jpg b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s6.jpg new file mode 100644 index 0000000000000000000000000000000000000000..91c13d5f2b48d143ca596566ad10f0a0e5693da4 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s6.jpg differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s7.jpg b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s7.jpg new file mode 100644 index 0000000000000000000000000000000000000000..cf96f2d5651f7ae0faf08193ecd3df282c5c3b53 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s7.jpg differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s8.jpg b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s8.jpg new file mode 100644 index 0000000000000000000000000000000000000000..b415ed1d4a4e5cf01e6dc30d6b4ced20814558d5 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s8.jpg differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s9.jpg b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s9.jpg new file mode 100644 index 0000000000000000000000000000000000000000..3ef7251ba10bf83356587016b126a52bdbca7b18 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/original_sample_asset/source/s9.jpg differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/sample_exp_data.exp b/ComfyUI-AdvancedLivePortrait/sample/sample_exp_data.exp new file mode 100644 index 0000000000000000000000000000000000000000..66c75aaa3ebcabbab5f49ac3c3663898f5f0e1d7 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/sample_exp_data.exp differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/source_image.png b/ComfyUI-AdvancedLivePortrait/sample/source_image.png new file mode 100644 index 0000000000000000000000000000000000000000..bbc3a122eb9d48b342e1412948c6e1331f50c1ea Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/source_image.png differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/workflow_advanced_screenshot.png b/ComfyUI-AdvancedLivePortrait/sample/workflow_advanced_screenshot.png new file mode 100644 index 0000000000000000000000000000000000000000..755b0cb1a22ed6860fe3d90e4278b60eda6b6b13 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/workflow_advanced_screenshot.png differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/workflow_basic_screenshot.png b/ComfyUI-AdvancedLivePortrait/sample/workflow_basic_screenshot.png new file mode 100644 index 0000000000000000000000000000000000000000..6894fa2c2d049bf61aed6f0686c437f92ed813d4 Binary files /dev/null and b/ComfyUI-AdvancedLivePortrait/sample/workflow_basic_screenshot.png differ diff --git a/ComfyUI-AdvancedLivePortrait/sample/workflows/Advanced_Animate_without_vid.json b/ComfyUI-AdvancedLivePortrait/sample/workflows/Advanced_Animate_without_vid.json new file mode 100644 index 0000000000000000000000000000000000000000..8edd32dba454d9f1489760a7963e27cc19203d0e --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/sample/workflows/Advanced_Animate_without_vid.json @@ -0,0 +1,359 @@ +{ + "last_node_id": 18, + "last_link_id": 11, + "nodes": [ + { + "id": 3, + "type": "LoadImage", + "pos": [ + 1385, + 278 + ], + "size": [ + 223.20395091844193, + 308.6326788490376 + ], + "flags": {}, + "order": 0, + "mode": 0, + "outputs": [ + { + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 11 + ], + "slot_index": 0, + "shape": 3 + }, + { + "name": "MASK", + "type": "MASK", + "links": null, + "slot_index": 1, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "LoadImage" + }, + "widgets_values": [ + "source_image.png", + "image" + ] + }, + { + "id": 7, + "type": "AdvancedLivePortrait", + "pos": [ + 1889, + 270 + ], + "size": { + "0": 235.1999969482422, + "1": 523.0364379882812 + }, + "flags": {}, + "order": 5, + "mode": 0, + "inputs": [ + { + "name": "src_images", + "type": "IMAGE", + "link": null + }, + { + "name": "motion_link", + "type": "EDITOR_LINK", + "link": 10 + }, + { + "name": "driving_images", + "type": "IMAGE", + "link": null + } + ], + "outputs": [ + { + "name": "images", + "type": "IMAGE", + "links": [ + 8 + ], + "slot_index": 0, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "AdvancedLivePortrait" + }, + "widgets_values": [ + 0, + 0, + 1.7000000000000002, + true, + false, + true, + "1 = 5:5\n0 = 3:5" + ] + }, + { + "id": 10, + "type": "VHS_VideoCombine", + "pos": [ + 2163, + 203 + ], + "size": [ + 261.79998779296875, + 565.7999877929688 + ], + "flags": {}, + "order": 6, + "mode": 0, + "inputs": [ + { + "name": "images", + "type": "IMAGE", + "link": 8 + }, + { + "name": "audio", + "type": "AUDIO", + "link": null + }, + { + "name": "meta_batch", + "type": "VHS_BatchManager", + "link": null + }, + { + "name": "vae", + "type": "VAE", + "link": null + } + ], + "outputs": [ + { + "name": "Filenames", + "type": "VHS_FILENAMES", + "links": null, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "VHS_VideoCombine" + }, + "widgets_values": { + "frame_rate": 15, + "loop_count": 0, + "filename_prefix": "AdvancedLivePortrait", + "format": "video/h264-mp4", + "pix_fmt": "yuv420p", + "crf": 19, + "save_metadata": true, + "pingpong": false, + "save_output": true, + "videopreview": { + "hidden": false, + "paused": false, + "params": { + "filename": "AdvancedLivePortrait_00021.mp4", + "subfolder": "", + "type": "output", + "format": "video/h264-mp4", + "frame_rate": 15 + } + } + } + }, + { + "id": 15, + "type": "ExpressionEditor", + "pos": [ + 1643, + 271 + ], + "size": { + "0": 210, + "1": 690 + }, + "flags": {}, + "order": 4, + "mode": 0, + "inputs": [ + { + "name": "src_image", + "type": "IMAGE", + "link": 11 + }, + { + "name": "motion_link", + "type": "EDITOR_LINK", + "link": null + }, + { + "name": "sample_image", + "type": "IMAGE", + "link": null + }, + { + "name": "add_exp", + "type": "EXP_DATA", + "link": null + } + ], + "outputs": [ + { + "name": "image", + "type": "IMAGE", + "links": null, + "shape": 3 + }, + { + "name": "motion_link", + "type": "EDITOR_LINK", + "links": [ + 10 + ], + "slot_index": 1, + "shape": 3 + }, + { + "name": "save_exp", + "type": "EXP_DATA", + "links": null, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "ExpressionEditor" + }, + "widgets_values": [ + 0, + 0, + 0, + 0, + 0, + 25, + 0, + 0, + 0, + 0, + 0, + 0, + 1, + 1, + "OnlyExpression", + 1.7 + ] + }, + { + "id": 16, + "type": "Note", + "pos": [ + 1892, + 831 + ], + "size": [ + 483.8180073265835, + 161.2032413725933 + ], + "flags": {}, + "order": 3, + "mode": 0, + "properties": { + "text": "" + }, + "widgets_values": [ + "Command Description\n1 = 5:5\n0 = 3:5\n\n[Motion index] = [Changing frame length] : [Length of frames waiting for next motion]\n\nfor example, first line -> 1 = 5:5\nChange over 5 frames with motion 1, and wait for the next motion for 5 frames" + ], + "color": "#432", + "bgcolor": "#653" + }, + { + "id": 17, + "type": "Note", + "pos": [ + 1629, + 996 + ], + "size": [ + 227.12395091844178, + 58 + ], + "flags": {}, + "order": 2, + "mode": 0, + "properties": { + "text": "" + }, + "widgets_values": [ + "motion 1" + ], + "color": "#432", + "bgcolor": "#653" + }, + { + "id": 18, + "type": "Note", + "pos": [ + 1373, + 625 + ], + "size": { + "0": 210, + "1": 58 + }, + "flags": {}, + "order": 1, + "mode": 0, + "properties": { + "text": "" + }, + "widgets_values": [ + "orginal src : motion 0\n" + ], + "color": "#432", + "bgcolor": "#653" + } + ], + "links": [ + [ + 8, + 7, + 0, + 10, + 0, + "IMAGE" + ], + [ + 10, + 15, + 1, + 7, + 1, + "EDITOR_LINK" + ], + [ + 11, + 3, + 0, + 15, + 0, + "IMAGE" + ] + ], + "groups": [], + "config": {}, + "extra": { + "ds": { + "scale": 1.1000000000000005, + "offset": [ + -841.7371241966999, + 104.19308875020204 + ] + } + }, + "version": 0.4 +} \ No newline at end of file diff --git a/ComfyUI-AdvancedLivePortrait/sample/workflows/Advanced_Inserting_expressions_into_vid.json b/ComfyUI-AdvancedLivePortrait/sample/workflows/Advanced_Inserting_expressions_into_vid.json new file mode 100644 index 0000000000000000000000000000000000000000..cad2fa439c20747cd571f34188e960d6aaa2993e --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/sample/workflows/Advanced_Inserting_expressions_into_vid.json @@ -0,0 +1,564 @@ +{ + "last_node_id": 14, + "last_link_id": 8, + "nodes": [ + { + "id": 3, + "type": "LoadImage", + "pos": [ + 736, + 299 + ], + "size": { + "0": 315, + "1": 314 + }, + "flags": {}, + "order": 0, + "mode": 0, + "outputs": [ + { + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 3 + ], + "slot_index": 0, + "shape": 3 + }, + { + "name": "MASK", + "type": "MASK", + "links": null, + "slot_index": 1, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "LoadImage" + }, + "widgets_values": [ + "source_image.png", + "image" + ] + }, + { + "id": 6, + "type": "ExpressionEditor", + "pos": [ + 1085, + 294 + ], + "size": { + "0": 210, + "1": 690 + }, + "flags": {}, + "order": 6, + "mode": 0, + "inputs": [ + { + "name": "src_image", + "type": "IMAGE", + "link": 3 + }, + { + "name": "motion_link", + "type": "EDITOR_LINK", + "link": null + }, + { + "name": "sample_image", + "type": "IMAGE", + "link": null + }, + { + "name": "add_exp", + "type": "EXP_DATA", + "link": null + } + ], + "outputs": [ + { + "name": "image", + "type": "IMAGE", + "links": null, + "shape": 3 + }, + { + "name": "motion_link", + "type": "EDITOR_LINK", + "links": [ + 5 + ], + "slot_index": 1, + "shape": 3 + }, + { + "name": "save_exp", + "type": "EXP_DATA", + "links": null, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "ExpressionEditor" + }, + "widgets_values": [ + 0, + 0, + 0, + 0, + 0, + 0, + -15, + 0, + 0, + 0, + 0, + 0, + 1, + 1, + "OnlyExpression", + 1.7000000000000002 + ] + }, + { + "id": 7, + "type": "AdvancedLivePortrait", + "pos": [ + 1845, + 304 + ], + "size": { + "0": 235.1999969482422, + "1": 523.0364379882812 + }, + "flags": {}, + "order": 8, + "mode": 0, + "inputs": [ + { + "name": "src_images", + "type": "IMAGE", + "link": null + }, + { + "name": "motion_link", + "type": "EDITOR_LINK", + "link": 6 + }, + { + "name": "driving_images", + "type": "IMAGE", + "link": 7 + } + ], + "outputs": [ + { + "name": "images", + "type": "IMAGE", + "links": [ + 8 + ], + "slot_index": 0, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "AdvancedLivePortrait" + }, + "widgets_values": [ + 0, + 0, + 1.7000000000000002, + "1 = 1:10\n2 = 5:10\n0 = 2:50\n1 = 2:0", + "1 = 1:10\n2 = 5:10\n0 = 2:50\n1 = 2:0", + false, + "1 = 1:10\n2 = 5:10\n0 = 2:50\n1 = 2:0" + ] + }, + { + "id": 8, + "type": "ExpressionEditor", + "pos": [ + 1342, + 304 + ], + "size": { + "0": 210, + "1": 690 + }, + "flags": {}, + "order": 7, + "mode": 0, + "inputs": [ + { + "name": "src_image", + "type": "IMAGE", + "link": null + }, + { + "name": "motion_link", + "type": "EDITOR_LINK", + "link": 5 + }, + { + "name": "sample_image", + "type": "IMAGE", + "link": null + }, + { + "name": "add_exp", + "type": "EXP_DATA", + "link": null + } + ], + "outputs": [ + { + "name": "image", + "type": "IMAGE", + "links": null, + "shape": 3 + }, + { + "name": "motion_link", + "type": "EDITOR_LINK", + "links": [ + 6 + ], + "slot_index": 1, + "shape": 3 + }, + { + "name": "save_exp", + "type": "EXP_DATA", + "links": null, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "ExpressionEditor" + }, + "widgets_values": [ + 0, + 0, + 0, + 0, + 0, + 0, + 15, + 0, + 0, + 0, + 0, + 0, + 1, + 1, + "OnlyExpression", + 1.7 + ] + }, + { + "id": 9, + "type": "VHS_LoadVideo", + "pos": [ + 1570, + 399 + ], + "size": [ + 235.1999969482422, + 491.1999969482422 + ], + "flags": {}, + "order": 1, + "mode": 0, + "inputs": [ + { + "name": "meta_batch", + "type": "VHS_BatchManager", + "link": null + }, + { + "name": "vae", + "type": "VAE", + "link": null + } + ], + "outputs": [ + { + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 7 + ], + "slot_index": 0, + "shape": 3 + }, + { + "name": "frame_count", + "type": "INT", + "links": null, + "shape": 3 + }, + { + "name": "audio", + "type": "AUDIO", + "links": null, + "shape": 3 + }, + { + "name": "video_info", + "type": "VHS_VIDEOINFO", + "links": null, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "VHS_LoadVideo" + }, + "widgets_values": { + "video": "driving_video.mp4", + "force_rate": 0, + "force_size": "Disabled", + "custom_width": 512, + "custom_height": 512, + "frame_load_cap": 0, + "skip_first_frames": 0, + "select_every_nth": 2, + "choose video to upload": "image", + "videopreview": { + "hidden": false, + "paused": false, + "params": { + "frame_load_cap": 0, + "skip_first_frames": 0, + "force_rate": 0, + "filename": "driving_video.mp4", + "type": "input", + "format": "video/mp4", + "select_every_nth": 2 + } + } + } + }, + { + "id": 10, + "type": "VHS_VideoCombine", + "pos": [ + 2117, + 305 + ], + "size": [ + 261.79998779296875, + 565.7999877929688 + ], + "flags": {}, + "order": 9, + "mode": 0, + "inputs": [ + { + "name": "images", + "type": "IMAGE", + "link": 8 + }, + { + "name": "audio", + "type": "AUDIO", + "link": null + }, + { + "name": "meta_batch", + "type": "VHS_BatchManager", + "link": null + }, + { + "name": "vae", + "type": "VAE", + "link": null + } + ], + "outputs": [ + { + "name": "Filenames", + "type": "VHS_FILENAMES", + "links": null, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "VHS_VideoCombine" + }, + "widgets_values": { + "frame_rate": 15, + "loop_count": 0, + "filename_prefix": "AdvancedLivePortrait", + "format": "video/h264-mp4", + "pix_fmt": "yuv420p", + "crf": 19, + "save_metadata": true, + "pingpong": false, + "save_output": true, + "videopreview": { + "hidden": false, + "paused": false, + "params": { + "filename": "AdvancedLivePortrait_00023.mp4", + "subfolder": "", + "type": "output", + "format": "video/h264-mp4", + "frame_rate": 15 + } + } + } + }, + { + "id": 11, + "type": "Note", + "pos": [ + 1834, + 935 + ], + "size": { + "0": 539.0599975585938, + "1": 184.96444702148438 + }, + "flags": {}, + "order": 5, + "mode": 0, + "properties": { + "text": "" + }, + "widgets_values": [ + "Command Description\n1 = 1:10\n2 = 5:10\n0 = 2:50\n1 = 2:0\n\n[Motion index] = [Changing frame length] : [Length of frames waiting for next motion]\n\nfor example, second line -> 2 = 5:10\nChange over 5 frames with motion 2, and wait for the next motion for 10 frames\n\n" + ], + "color": "#432", + "bgcolor": "#653" + }, + { + "id": 12, + "type": "Note", + "pos": [ + 1089, + 1038 + ], + "size": { + "0": 210, + "1": 58 + }, + "flags": {}, + "order": 2, + "mode": 0, + "properties": { + "text": "" + }, + "widgets_values": [ + "motion1\n" + ], + "color": "#432", + "bgcolor": "#653" + }, + { + "id": 13, + "type": "Note", + "pos": [ + 1341, + 1041 + ], + "size": { + "0": 210, + "1": 58 + }, + "flags": { + "collapsed": false + }, + "order": 3, + "mode": 0, + "properties": { + "text": "" + }, + "widgets_values": [ + "motion2\n" + ], + "color": "#432", + "bgcolor": "#653" + }, + { + "id": 14, + "type": "Note", + "pos": [ + 776, + 684 + ], + "size": { + "0": 210, + "1": 58 + }, + "flags": {}, + "order": 4, + "mode": 0, + "properties": { + "text": "" + }, + "widgets_values": [ + "orginal src : motion0\n" + ], + "color": "#432", + "bgcolor": "#653" + } + ], + "links": [ + [ + 3, + 3, + 0, + 6, + 0, + "IMAGE" + ], + [ + 5, + 6, + 1, + 8, + 1, + "EDITOR_LINK" + ], + [ + 6, + 8, + 1, + 7, + 1, + "EDITOR_LINK" + ], + [ + 7, + 9, + 0, + 7, + 2, + "IMAGE" + ], + [ + 8, + 7, + 0, + 10, + 0, + "IMAGE" + ] + ], + "groups": [], + "config": {}, + "extra": { + "ds": { + "scale": 1.1000000000000005, + "offset": [ + -678.8740830330757, + -35.741089089667945 + ] + } + }, + "version": 0.4 +} \ No newline at end of file diff --git a/ComfyUI-AdvancedLivePortrait/sample/workflows/Bacis_Extracting_expression_from_photo.json b/ComfyUI-AdvancedLivePortrait/sample/workflows/Bacis_Extracting_expression_from_photo.json new file mode 100644 index 0000000000000000000000000000000000000000..bb305e611c00e3b7ff5468dfd186208cca36fba6 --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/sample/workflows/Bacis_Extracting_expression_from_photo.json @@ -0,0 +1,241 @@ +{ + "last_node_id": 33, + "last_link_id": 39, + "nodes": [ + { + "id": 19, + "type": "LoadImage", + "pos": [ + 1022, + 702 + ], + "size": { + "0": 220.3717041015625, + "1": 314 + }, + "flags": {}, + "order": 1, + "mode": 0, + "outputs": [ + { + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 25 + ], + "slot_index": 0, + "shape": 3 + }, + { + "name": "MASK", + "type": "MASK", + "links": null, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "LoadImage" + }, + "widgets_values": [ + "exp_image.png", + "image" + ] + }, + { + "id": 20, + "type": "LoadImage", + "pos": [ + 1027.712857031252, + 322.70146904296865 + ], + "size": { + "0": 210.9286651611328, + "1": 314 + }, + "flags": {}, + "order": 0, + "mode": 0, + "outputs": [ + { + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 26 + ], + "slot_index": 0, + "shape": 3 + }, + { + "name": "MASK", + "type": "MASK", + "links": null, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "LoadImage" + }, + "widgets_values": [ + "source_image.png", + "image" + ] + }, + { + "id": 21, + "type": "ExpressionEditor", + "pos": [ + 1305, + 338 + ], + "size": [ + 252.62867736816406, + 690 + ], + "flags": {}, + "order": 2, + "mode": 0, + "inputs": [ + { + "name": "src_image", + "type": "IMAGE", + "link": 26 + }, + { + "name": "motion_link", + "type": "EDITOR_LINK", + "link": null + }, + { + "name": "sample_image", + "type": "IMAGE", + "link": 25 + }, + { + "name": "add_exp", + "type": "EXP_DATA", + "link": null + } + ], + "outputs": [ + { + "name": "image", + "type": "IMAGE", + "links": [ + 39 + ], + "slot_index": 0, + "shape": 3 + }, + { + "name": "motion_link", + "type": "EDITOR_LINK", + "links": null, + "shape": 3 + }, + { + "name": "save_exp", + "type": "EXP_DATA", + "links": null, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "ExpressionEditor" + }, + "widgets_values": [ + 0, + 0, + 0, + 0, + 0, + 0, + 0, + 0, + 0, + 0, + 0, + 0, + 1, + 1, + "OnlyExpression", + 1.7 + ] + }, + { + "id": 33, + "type": "SaveImage", + "pos": [ + 1617, + 390 + ], + "size": [ + 263.6363636363635, + 278.18181818181813 + ], + "flags": {}, + "order": 3, + "mode": 0, + "inputs": [ + { + "name": "images", + "type": "IMAGE", + "link": 39 + } + ], + "properties": {}, + "widgets_values": [ + "ExpressionEditing" + ] + } + ], + "links": [ + [ + 25, + 19, + 0, + 21, + 2, + "IMAGE" + ], + [ + 26, + 20, + 0, + 21, + 0, + "IMAGE" + ], + [ + 39, + 21, + 0, + 33, + 0, + "IMAGE" + ] + ], + "groups": [ + { + "title": "Extracting facial expressions from photos", + "bounding": [ + 970, + 233, + 939, + 823 + ], + "color": "#a1309b", + "font_size": 30 + } + ], + "config": {}, + "extra": { + "ds": { + "scale": 1, + "offset": [ + -256.9678555487616, + -36.006961063159906 + ] + } + }, + "version": 0.4 +} \ No newline at end of file diff --git a/ComfyUI-AdvancedLivePortrait/sample/workflows/Bacis_Liveportrait_with_expression_editing.json b/ComfyUI-AdvancedLivePortrait/sample/workflows/Bacis_Liveportrait_with_expression_editing.json new file mode 100644 index 0000000000000000000000000000000000000000..031b10342e4adf6525f5335d86a8d3b13b7dc10e --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/sample/workflows/Bacis_Liveportrait_with_expression_editing.json @@ -0,0 +1,392 @@ +{ + "last_node_id": 32, + "last_link_id": 38, + "nodes": [ + { + "id": 26, + "type": "LoadImage", + "pos": [ + 574.0879504882818, + 354.13510488281264 + ], + "size": { + "0": 210, + "1": 314 + }, + "flags": {}, + "order": 0, + "mode": 0, + "outputs": [ + { + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 30 + ], + "slot_index": 0, + "shape": 3 + }, + { + "name": "MASK", + "type": "MASK", + "links": null, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "LoadImage" + }, + "widgets_values": [ + "source_image.png", + "image" + ] + }, + { + "id": 27, + "type": "VHS_LoadVideo", + "pos": [ + 1113, + 443 + ], + "size": [ + 235.1999969482422, + 491.1999969482422 + ], + "flags": {}, + "order": 1, + "mode": 0, + "inputs": [ + { + "name": "meta_batch", + "type": "VHS_BatchManager", + "link": null + }, + { + "name": "vae", + "type": "VAE", + "link": null + } + ], + "outputs": [ + { + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 31 + ], + "slot_index": 0, + "shape": 3 + }, + { + "name": "frame_count", + "type": "INT", + "links": null, + "shape": 3 + }, + { + "name": "audio", + "type": "AUDIO", + "links": null, + "shape": 3 + }, + { + "name": "video_info", + "type": "VHS_VIDEOINFO", + "links": null, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "VHS_LoadVideo" + }, + "widgets_values": { + "video": "driving_video.mp4", + "force_rate": 0, + "force_size": "Disabled", + "custom_width": 512, + "custom_height": 512, + "frame_load_cap": 0, + "skip_first_frames": 0, + "select_every_nth": 2, + "choose video to upload": "image", + "videopreview": { + "hidden": false, + "paused": false, + "params": { + "frame_load_cap": 0, + "skip_first_frames": 0, + "force_rate": 0, + "filename": "driving_video.mp4", + "type": "input", + "format": "video/mp4", + "select_every_nth": 2 + } + } + } + }, + { + "id": 28, + "type": "AdvancedLivePortrait", + "pos": [ + 1386, + 353 + ], + "size": [ + 235.1999969482422, + 674.0390319824219 + ], + "flags": {}, + "order": 3, + "mode": 0, + "inputs": [ + { + "name": "src_images", + "type": "IMAGE", + "link": 38 + }, + { + "name": "motion_link", + "type": "EDITOR_LINK", + "link": null + }, + { + "name": "driving_images", + "type": "IMAGE", + "link": 31 + } + ], + "outputs": [ + { + "name": "images", + "type": "IMAGE", + "links": [ + 35 + ], + "slot_index": 0, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "AdvancedLivePortrait" + }, + "widgets_values": [ + 0, + 0, + 1.5, + true, + "", + false, + "" + ] + }, + { + "id": 29, + "type": "VHS_VideoCombine", + "pos": [ + 1664, + 357 + ], + "size": [ + 272.9300842285156, + 576.9300842285156 + ], + "flags": {}, + "order": 4, + "mode": 0, + "inputs": [ + { + "name": "images", + "type": "IMAGE", + "link": 35 + }, + { + "name": "audio", + "type": "AUDIO", + "link": null + }, + { + "name": "meta_batch", + "type": "VHS_BatchManager", + "link": null + }, + { + "name": "vae", + "type": "VAE", + "link": null + } + ], + "outputs": [ + { + "name": "Filenames", + "type": "VHS_FILENAMES", + "links": null, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "VHS_VideoCombine" + }, + "widgets_values": { + "frame_rate": 15, + "loop_count": 0, + "filename_prefix": "AdvancedLivePortrait", + "format": "video/h264-mp4", + "pix_fmt": "yuv420p", + "crf": 19, + "save_metadata": true, + "pingpong": false, + "save_output": true, + "videopreview": { + "hidden": false, + "paused": false, + "params": { + "filename": "AdvancedLivePortrait_00001.mp4", + "subfolder": "", + "type": "temp", + "format": "video/h264-mp4", + "frame_rate": 15 + } + } + } + }, + { + "id": 30, + "type": "ExpressionEditor", + "pos": [ + 837.0879504882818, + 349.1351048828126 + ], + "size": { + "0": 256.8715515136719, + "1": 690 + }, + "flags": {}, + "order": 2, + "mode": 0, + "inputs": [ + { + "name": "src_image", + "type": "IMAGE", + "link": 30 + }, + { + "name": "motion_link", + "type": "EDITOR_LINK", + "link": null + }, + { + "name": "sample_image", + "type": "IMAGE", + "link": null + }, + { + "name": "add_exp", + "type": "EXP_DATA", + "link": null + } + ], + "outputs": [ + { + "name": "image", + "type": "IMAGE", + "links": [ + 38 + ], + "slot_index": 0, + "shape": 3 + }, + { + "name": "motion_link", + "type": "EDITOR_LINK", + "links": null, + "shape": 3 + }, + { + "name": "save_exp", + "type": "EXP_DATA", + "links": null, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "ExpressionEditor" + }, + "widgets_values": [ + 0, + 0, + 0, + 0, + 0, + 0, + 0, + 0, + 0, + 0, + 0, + 1.3, + 1, + 1, + "OnlyExpression", + 1.7000000000000002 + ] + } + ], + "links": [ + [ + 30, + 26, + 0, + 30, + 0, + "IMAGE" + ], + [ + 31, + 27, + 0, + 28, + 2, + "IMAGE" + ], + [ + 35, + 28, + 0, + 29, + 0, + "IMAGE" + ], + [ + 38, + 30, + 0, + 28, + 0, + "IMAGE" + ] + ], + "groups": [ + { + "title": "Live Portrait with facial expression editing", + "bounding": [ + 555, + 266, + 1427, + 789 + ], + "color": "#3f789e", + "font_size": 30 + } + ], + "config": {}, + "extra": { + "ds": { + "scale": 0.8264462809917354, + "offset": [ + 185.1062862025528, + 195.77375741392547 + ] + } + }, + "version": 0.4 +} \ No newline at end of file diff --git a/ComfyUI-AdvancedLivePortrait/sample/workflows/Bacis_Simple_expression_editing.json b/ComfyUI-AdvancedLivePortrait/sample/workflows/Bacis_Simple_expression_editing.json new file mode 100644 index 0000000000000000000000000000000000000000..ee60a3a9e06f6be80527844d26fb0379847cbe64 --- /dev/null +++ b/ComfyUI-AdvancedLivePortrait/sample/workflows/Bacis_Simple_expression_editing.json @@ -0,0 +1,194 @@ +{ + "last_node_id": 33, + "last_link_id": 40, + "nodes": [ + { + "id": 14, + "type": "ExpressionEditor", + "pos": [ + 1044.0000000000002, + 362.78992797851555 + ], + "size": [ + 260.81048583984375, + 690 + ], + "flags": {}, + "order": 1, + "mode": 0, + "inputs": [ + { + "name": "src_image", + "type": "IMAGE", + "link": 19 + }, + { + "name": "motion_link", + "type": "EDITOR_LINK", + "link": null + }, + { + "name": "sample_image", + "type": "IMAGE", + "link": null + }, + { + "name": "add_exp", + "type": "EXP_DATA", + "link": null + } + ], + "outputs": [ + { + "name": "image", + "type": "IMAGE", + "links": [ + 40 + ], + "slot_index": 0, + "shape": 3 + }, + { + "name": "motion_link", + "type": "EDITOR_LINK", + "links": null, + "shape": 3 + }, + { + "name": "save_exp", + "type": "EXP_DATA", + "links": null, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "ExpressionEditor" + }, + "widgets_values": [ + 0, + 0, + 0, + 0, + 0, + 23.5, + 0, + 0, + 0, + 0, + 0, + 0, + 1, + 1, + "OnlyExpression", + 1.7000000000000002 + ] + }, + { + "id": 15, + "type": "LoadImage", + "pos": [ + 776, + 366.78992797851555 + ], + "size": { + "0": 210, + "1": 314 + }, + "flags": {}, + "order": 0, + "mode": 0, + "outputs": [ + { + "name": "IMAGE", + "type": "IMAGE", + "links": [ + 19 + ], + "slot_index": 0, + "shape": 3 + }, + { + "name": "MASK", + "type": "MASK", + "links": null, + "shape": 3 + } + ], + "properties": { + "Node name for S&R": "LoadImage" + }, + "widgets_values": [ + "source_image.png", + "image" + ] + }, + { + "id": 33, + "type": "SaveImage", + "pos": [ + 1351.0000000000005, + 373.78992797851555 + ], + "size": [ + 257.83631037110194, + 270 + ], + "flags": {}, + "order": 2, + "mode": 0, + "inputs": [ + { + "name": "images", + "type": "IMAGE", + "link": 40 + } + ], + "properties": {}, + "widgets_values": [ + "ExpressionEditing" + ] + } + ], + "links": [ + [ + 19, + 15, + 0, + 14, + 0, + "IMAGE" + ], + [ + 40, + 14, + 0, + 33, + 0, + "IMAGE" + ] + ], + "groups": [ + { + "title": "Simple facial expression editing", + "bounding": [ + 745, + 261, + 896, + 835 + ], + "color": "#88A", + "font_size": 30 + } + ], + "config": {}, + "extra": { + "ds": { + "scale": 0.7513148009015777, + "offset": [ + 561.7325573974937, + 267.43204960931945 + ] + } + }, + "version": 0.4 +} \ No newline at end of file diff --git a/ComfyUI-Impact-Pack/.github/workflows/publish.yml b/ComfyUI-Impact-Pack/.github/workflows/publish.yml new file mode 100644 index 0000000000000000000000000000000000000000..6e7201833e56c009e347731016b54e1c6d2254ab --- /dev/null +++ b/ComfyUI-Impact-Pack/.github/workflows/publish.yml @@ -0,0 +1,21 @@ +name: Publish to Comfy registry +on: + workflow_dispatch: + push: + branches: + - main + paths: + - "pyproject.toml" + +jobs: + publish-node: + name: Publish Custom Node to registry + runs-on: ubuntu-latest + steps: + - name: Check out code + uses: actions/checkout@v4 + - name: Publish Custom Node + uses: Comfy-Org/publish-node-action@main + with: + ## Add your own personal access token to your Github Repository secrets and reference it here. + personal_access_token: ${{ secrets.REGISTRY_ACCESS_TOKEN }} \ No newline at end of file diff --git a/ComfyUI-Impact-Pack/.gitignore b/ComfyUI-Impact-Pack/.gitignore new file mode 100644 index 0000000000000000000000000000000000000000..194a93afce6f0f3e9707d3ba68e3a3a26be82eaf --- /dev/null +++ b/ComfyUI-Impact-Pack/.gitignore @@ -0,0 +1,9 @@ +__pycache__ +*.ini +wildcards/** +.vscode/ +.idea/ +subpack +impact_subpack +*.txt +*.yaml diff --git a/ComfyUI-Impact-Pack/.gitmodules b/ComfyUI-Impact-Pack/.gitmodules new file mode 100644 index 0000000000000000000000000000000000000000..dbbc2e070a45bc6b72ba8492a28d9c29737ef044 --- /dev/null +++ b/ComfyUI-Impact-Pack/.gitmodules @@ -0,0 +1,3 @@ +[submodule "subpack"] + path = subpack + url = https://github.com/ltdrdata/ComfyUI-Impact-Subpack diff --git a/ComfyUI-Impact-Pack/LICENSE.txt b/ComfyUI-Impact-Pack/LICENSE.txt new file mode 100644 index 0000000000000000000000000000000000000000..3877ae0a7ff6f94ac222fd704e112723db776114 --- /dev/null +++ b/ComfyUI-Impact-Pack/LICENSE.txt @@ -0,0 +1,674 @@ + GNU GENERAL PUBLIC LICENSE + Version 3, 29 June 2007 + + Copyright (C) 2007 Free Software Foundation, Inc. + Everyone is permitted to copy and distribute verbatim copies + of this license document, but changing it is not allowed. + + Preamble + + The GNU General Public License is a free, copyleft license for +software and other kinds of works. + + The licenses for most software and other practical works are designed +to take away your freedom to share and change the works. By contrast, +the GNU General Public License is intended to guarantee your freedom to +share and change all versions of a program--to make sure it remains free +software for all its users. We, the Free Software Foundation, use the +GNU General Public License for most of our software; it applies also to +any other work released this way by its authors. You can apply it to +your programs, too. + + When we speak of free software, we are referring to freedom, not +price. Our General Public Licenses are designed to make sure that you +have the freedom to distribute copies of free software (and charge for +them if you wish), that you receive source code or can get it if you +want it, that you can change the software or use pieces of it in new +free programs, and that you know you can do these things. + + To protect your rights, we need to prevent others from denying you +these rights or asking you to surrender the rights. Therefore, you have +certain responsibilities if you distribute copies of the software, or if +you modify it: responsibilities to respect the freedom of others. + + For example, if you distribute copies of such a program, whether +gratis or for a fee, you must pass on to the recipients the same +freedoms that you received. You must make sure that they, too, receive +or can get the source code. And you must show them these terms so they +know their rights. + + Developers that use the GNU GPL protect your rights with two steps: +(1) assert copyright on the software, and (2) offer you this License +giving you legal permission to copy, distribute and/or modify it. + + For the developers' and authors' protection, the GPL clearly explains +that there is no warranty for this free software. For both users' and +authors' sake, the GPL requires that modified versions be marked as +changed, so that their problems will not be attributed erroneously to +authors of previous versions. + + Some devices are designed to deny users access to install or run +modified versions of the software inside them, although the manufacturer +can do so. This is fundamentally incompatible with the aim of +protecting users' freedom to change the software. The systematic +pattern of such abuse occurs in the area of products for individuals to +use, which is precisely where it is most unacceptable. Therefore, we +have designed this version of the GPL to prohibit the practice for those +products. If such problems arise substantially in other domains, we +stand ready to extend this provision to those domains in future versions +of the GPL, as needed to protect the freedom of users. + + Finally, every program is threatened constantly by software patents. +States should not allow patents to restrict development and use of +software on general-purpose computers, but in those that do, we wish to +avoid the special danger that patents applied to a free program could +make it effectively proprietary. To prevent this, the GPL assures that +patents cannot be used to render the program non-free. + + The precise terms and conditions for copying, distribution and +modification follow. + + TERMS AND CONDITIONS + + 0. Definitions. + + "This License" refers to version 3 of the GNU General Public License. + + "Copyright" also means copyright-like laws that apply to other kinds of +works, such as semiconductor masks. + + "The Program" refers to any copyrightable work licensed under this +License. Each licensee is addressed as "you". "Licensees" and +"recipients" may be individuals or organizations. + + To "modify" a work means to copy from or adapt all or part of the work +in a fashion requiring copyright permission, other than the making of an +exact copy. The resulting work is called a "modified version" of the +earlier work or a work "based on" the earlier work. + + A "covered work" means either the unmodified Program or a work based +on the Program. + + To "propagate" a work means to do anything with it that, without +permission, would make you directly or secondarily liable for +infringement under applicable copyright law, except executing it on a +computer or modifying a private copy. Propagation includes copying, +distribution (with or without modification), making available to the +public, and in some countries other activities as well. + + To "convey" a work means any kind of propagation that enables other +parties to make or receive copies. Mere interaction with a user through +a computer network, with no transfer of a copy, is not conveying. + + An interactive user interface displays "Appropriate Legal Notices" +to the extent that it includes a convenient and prominently visible +feature that (1) displays an appropriate copyright notice, and (2) +tells the user that there is no warranty for the work (except to the +extent that warranties are provided), that licensees may convey the +work under this License, and how to view a copy of this License. If +the interface presents a list of user commands or options, such as a +menu, a prominent item in the list meets this criterion. + + 1. Source Code. + + The "source code" for a work means the preferred form of the work +for making modifications to it. "Object code" means any non-source +form of a work. + + A "Standard Interface" means an interface that either is an official +standard defined by a recognized standards body, or, in the case of +interfaces specified for a particular programming language, one that +is widely used among developers working in that language. + + The "System Libraries" of an executable work include anything, other +than the work as a whole, that (a) is included in the normal form of +packaging a Major Component, but which is not part of that Major +Component, and (b) serves only to enable use of the work with that +Major Component, or to implement a Standard Interface for which an +implementation is available to the public in source code form. A +"Major Component", in this context, means a major essential component +(kernel, window system, and so on) of the specific operating system +(if any) on which the executable work runs, or a compiler used to +produce the work, or an object code interpreter used to run it. + + The "Corresponding Source" for a work in object code form means all +the source code needed to generate, install, and (for an executable +work) run the object code and to modify the work, including scripts to +control those activities. However, it does not include the work's +System Libraries, or general-purpose tools or generally available free +programs which are used unmodified in performing those activities but +which are not part of the work. For example, Corresponding Source +includes interface definition files associated with source files for +the work, and the source code for shared libraries and dynamically +linked subprograms that the work is specifically designed to require, +such as by intimate data communication or control flow between those +subprograms and other parts of the work. + + The Corresponding Source need not include anything that users +can regenerate automatically from other parts of the Corresponding +Source. + + The Corresponding Source for a work in source code form is that +same work. + + 2. Basic Permissions. + + All rights granted under this License are granted for the term of +copyright on the Program, and are irrevocable provided the stated +conditions are met. This License explicitly affirms your unlimited +permission to run the unmodified Program. The output from running a +covered work is covered by this License only if the output, given its +content, constitutes a covered work. This License acknowledges your +rights of fair use or other equivalent, as provided by copyright law. + + You may make, run and propagate covered works that you do not +convey, without conditions so long as your license otherwise remains +in force. You may convey covered works to others for the sole purpose +of having them make modifications exclusively for you, or provide you +with facilities for running those works, provided that you comply with +the terms of this License in conveying all material for which you do +not control copyright. Those thus making or running the covered works +for you must do so exclusively on your behalf, under your direction +and control, on terms that prohibit them from making any copies of +your copyrighted material outside their relationship with you. + + Conveying under any other circumstances is permitted solely under +the conditions stated below. Sublicensing is not allowed; section 10 +makes it unnecessary. + + 3. Protecting Users' Legal Rights From Anti-Circumvention Law. + + No covered work shall be deemed part of an effective technological +measure under any applicable law fulfilling obligations under article +11 of the WIPO copyright treaty adopted on 20 December 1996, or +similar laws prohibiting or restricting circumvention of such +measures. + + When you convey a covered work, you waive any legal power to forbid +circumvention of technological measures to the extent such circumvention +is effected by exercising rights under this License with respect to +the covered work, and you disclaim any intention to limit operation or +modification of the work as a means of enforcing, against the work's +users, your or third parties' legal rights to forbid circumvention of +technological measures. + + 4. Conveying Verbatim Copies. + + You may convey verbatim copies of the Program's source code as you +receive it, in any medium, provided that you conspicuously and +appropriately publish on each copy an appropriate copyright notice; +keep intact all notices stating that this License and any +non-permissive terms added in accord with section 7 apply to the code; +keep intact all notices of the absence of any warranty; and give all +recipients a copy of this License along with the Program. + + You may charge any price or no price for each copy that you convey, +and you may offer support or warranty protection for a fee. + + 5. Conveying Modified Source Versions. + + You may convey a work based on the Program, or the modifications to +produce it from the Program, in the form of source code under the +terms of section 4, provided that you also meet all of these conditions: + + a) The work must carry prominent notices stating that you modified + it, and giving a relevant date. + + b) The work must carry prominent notices stating that it is + released under this License and any conditions added under section + 7. This requirement modifies the requirement in section 4 to + "keep intact all notices". + + c) You must license the entire work, as a whole, under this + License to anyone who comes into possession of a copy. This + License will therefore apply, along with any applicable section 7 + additional terms, to the whole of the work, and all its parts, + regardless of how they are packaged. This License gives no + permission to license the work in any other way, but it does not + invalidate such permission if you have separately received it. + + d) If the work has interactive user interfaces, each must display + Appropriate Legal Notices; however, if the Program has interactive + interfaces that do not display Appropriate Legal Notices, your + work need not make them do so. + + A compilation of a covered work with other separate and independent +works, which are not by their nature extensions of the covered work, +and which are not combined with it such as to form a larger program, +in or on a volume of a storage or distribution medium, is called an +"aggregate" if the compilation and its resulting copyright are not +used to limit the access or legal rights of the compilation's users +beyond what the individual works permit. Inclusion of a covered work +in an aggregate does not cause this License to apply to the other +parts of the aggregate. + + 6. Conveying Non-Source Forms. + + You may convey a covered work in object code form under the terms +of sections 4 and 5, provided that you also convey the +machine-readable Corresponding Source under the terms of this License, +in one of these ways: + + a) Convey the object code in, or embodied in, a physical product + (including a physical distribution medium), accompanied by the + Corresponding Source fixed on a durable physical medium + customarily used for software interchange. + + b) Convey the object code in, or embodied in, a physical product + (including a physical distribution medium), accompanied by a + written offer, valid for at least three years and valid for as + long as you offer spare parts or customer support for that product + model, to give anyone who possesses the object code either (1) a + copy of the Corresponding Source for all the software in the + product that is covered by this License, on a durable physical + medium customarily used for software interchange, for a price no + more than your reasonable cost of physically performing this + conveying of source, or (2) access to copy the + Corresponding Source from a network server at no charge. + + c) Convey individual copies of the object code with a copy of the + written offer to provide the Corresponding Source. This + alternative is allowed only occasionally and noncommercially, and + only if you received the object code with such an offer, in accord + with subsection 6b. + + d) Convey the object code by offering access from a designated + place (gratis or for a charge), and offer equivalent access to the + Corresponding Source in the same way through the same place at no + further charge. You need not require recipients to copy the + Corresponding Source along with the object code. If the place to + copy the object code is a network server, the Corresponding Source + may be on a different server (operated by you or a third party) + that supports equivalent copying facilities, provided you maintain + clear directions next to the object code saying where to find the + Corresponding Source. Regardless of what server hosts the + Corresponding Source, you remain obligated to ensure that it is + available for as long as needed to satisfy these requirements. + + e) Convey the object code using peer-to-peer transmission, provided + you inform other peers where the object code and Corresponding + Source of the work are being offered to the general public at no + charge under subsection 6d. + + A separable portion of the object code, whose source code is excluded +from the Corresponding Source as a System Library, need not be +included in conveying the object code work. + + A "User Product" is either (1) a "consumer product", which means any +tangible personal property which is normally used for personal, family, +or household purposes, or (2) anything designed or sold for incorporation +into a dwelling. In determining whether a product is a consumer product, +doubtful cases shall be resolved in favor of coverage. For a particular +product received by a particular user, "normally used" refers to a +typical or common use of that class of product, regardless of the status +of the particular user or of the way in which the particular user +actually uses, or expects or is expected to use, the product. A product +is a consumer product regardless of whether the product has substantial +commercial, industrial or non-consumer uses, unless such uses represent +the only significant mode of use of the product. + + "Installation Information" for a User Product means any methods, +procedures, authorization keys, or other information required to install +and execute modified versions of a covered work in that User Product from +a modified version of its Corresponding Source. The information must +suffice to ensure that the continued functioning of the modified object +code is in no case prevented or interfered with solely because +modification has been made. + + If you convey an object code work under this section in, or with, or +specifically for use in, a User Product, and the conveying occurs as +part of a transaction in which the right of possession and use of the +User Product is transferred to the recipient in perpetuity or for a +fixed term (regardless of how the transaction is characterized), the +Corresponding Source conveyed under this section must be accompanied +by the Installation Information. But this requirement does not apply +if neither you nor any third party retains the ability to install +modified object code on the User Product (for example, the work has +been installed in ROM). + + The requirement to provide Installation Information does not include a +requirement to continue to provide support service, warranty, or updates +for a work that has been modified or installed by the recipient, or for +the User Product in which it has been modified or installed. Access to a +network may be denied when the modification itself materially and +adversely affects the operation of the network or violates the rules and +protocols for communication across the network. + + Corresponding Source conveyed, and Installation Information provided, +in accord with this section must be in a format that is publicly +documented (and with an implementation available to the public in +source code form), and must require no special password or key for +unpacking, reading or copying. + + 7. Additional Terms. + + "Additional permissions" are terms that supplement the terms of this +License by making exceptions from one or more of its conditions. +Additional permissions that are applicable to the entire Program shall +be treated as though they were included in this License, to the extent +that they are valid under applicable law. If additional permissions +apply only to part of the Program, that part may be used separately +under those permissions, but the entire Program remains governed by +this License without regard to the additional permissions. + + When you convey a copy of a covered work, you may at your option +remove any additional permissions from that copy, or from any part of +it. (Additional permissions may be written to require their own +removal in certain cases when you modify the work.) You may place +additional permissions on material, added by you to a covered work, +for which you have or can give appropriate copyright permission. + + Notwithstanding any other provision of this License, for material you +add to a covered work, you may (if authorized by the copyright holders of +that material) supplement the terms of this License with terms: + + a) Disclaiming warranty or limiting liability differently from the + terms of sections 15 and 16 of this License; or + + b) Requiring preservation of specified reasonable legal notices or + author attributions in that material or in the Appropriate Legal + Notices displayed by works containing it; or + + c) Prohibiting misrepresentation of the origin of that material, or + requiring that modified versions of such material be marked in + reasonable ways as different from the original version; or + + d) Limiting the use for publicity purposes of names of licensors or + authors of the material; or + + e) Declining to grant rights under trademark law for use of some + trade names, trademarks, or service marks; or + + f) Requiring indemnification of licensors and authors of that + material by anyone who conveys the material (or modified versions of + it) with contractual assumptions of liability to the recipient, for + any liability that these contractual assumptions directly impose on + those licensors and authors. + + All other non-permissive additional terms are considered "further +restrictions" within the meaning of section 10. If the Program as you +received it, or any part of it, contains a notice stating that it is +governed by this License along with a term that is a further +restriction, you may remove that term. If a license document contains +a further restriction but permits relicensing or conveying under this +License, you may add to a covered work material governed by the terms +of that license document, provided that the further restriction does +not survive such relicensing or conveying. + + If you add terms to a covered work in accord with this section, you +must place, in the relevant source files, a statement of the +additional terms that apply to those files, or a notice indicating +where to find the applicable terms. + + Additional terms, permissive or non-permissive, may be stated in the +form of a separately written license, or stated as exceptions; +the above requirements apply either way. + + 8. Termination. + + You may not propagate or modify a covered work except as expressly +provided under this License. Any attempt otherwise to propagate or +modify it is void, and will automatically terminate your rights under +this License (including any patent licenses granted under the third +paragraph of section 11). + + However, if you cease all violation of this License, then your +license from a particular copyright holder is reinstated (a) +provisionally, unless and until the copyright holder explicitly and +finally terminates your license, and (b) permanently, if the copyright +holder fails to notify you of the violation by some reasonable means +prior to 60 days after the cessation. + + Moreover, your license from a particular copyright holder is +reinstated permanently if the copyright holder notifies you of the +violation by some reasonable means, this is the first time you have +received notice of violation of this License (for any work) from that +copyright holder, and you cure the violation prior to 30 days after +your receipt of the notice. + + Termination of your rights under this section does not terminate the +licenses of parties who have received copies or rights from you under +this License. If your rights have been terminated and not permanently +reinstated, you do not qualify to receive new licenses for the same +material under section 10. + + 9. Acceptance Not Required for Having Copies. + + You are not required to accept this License in order to receive or +run a copy of the Program. Ancillary propagation of a covered work +occurring solely as a consequence of using peer-to-peer transmission +to receive a copy likewise does not require acceptance. However, +nothing other than this License grants you permission to propagate or +modify any covered work. These actions infringe copyright if you do +not accept this License. Therefore, by modifying or propagating a +covered work, you indicate your acceptance of this License to do so. + + 10. Automatic Licensing of Downstream Recipients. + + Each time you convey a covered work, the recipient automatically +receives a license from the original licensors, to run, modify and +propagate that work, subject to this License. You are not responsible +for enforcing compliance by third parties with this License. + + An "entity transaction" is a transaction transferring control of an +organization, or substantially all assets of one, or subdividing an +organization, or merging organizations. If propagation of a covered +work results from an entity transaction, each party to that +transaction who receives a copy of the work also receives whatever +licenses to the work the party's predecessor in interest had or could +give under the previous paragraph, plus a right to possession of the +Corresponding Source of the work from the predecessor in interest, if +the predecessor has it or can get it with reasonable efforts. + + You may not impose any further restrictions on the exercise of the +rights granted or affirmed under this License. For example, you may +not impose a license fee, royalty, or other charge for exercise of +rights granted under this License, and you may not initiate litigation +(including a cross-claim or counterclaim in a lawsuit) alleging that +any patent claim is infringed by making, using, selling, offering for +sale, or importing the Program or any portion of it. + + 11. Patents. + + A "contributor" is a copyright holder who authorizes use under this +License of the Program or a work on which the Program is based. The +work thus licensed is called the contributor's "contributor version". + + A contributor's "essential patent claims" are all patent claims +owned or controlled by the contributor, whether already acquired or +hereafter acquired, that would be infringed by some manner, permitted +by this License, of making, using, or selling its contributor version, +but do not include claims that would be infringed only as a +consequence of further modification of the contributor version. For +purposes of this definition, "control" includes the right to grant +patent sublicenses in a manner consistent with the requirements of +this License. + + Each contributor grants you a non-exclusive, worldwide, royalty-free +patent license under the contributor's essential patent claims, to +make, use, sell, offer for sale, import and otherwise run, modify and +propagate the contents of its contributor version. + + In the following three paragraphs, a "patent license" is any express +agreement or commitment, however denominated, not to enforce a patent +(such as an express permission to practice a patent or covenant not to +sue for patent infringement). To "grant" such a patent license to a +party means to make such an agreement or commitment not to enforce a +patent against the party. + + If you convey a covered work, knowingly relying on a patent license, +and the Corresponding Source of the work is not available for anyone +to copy, free of charge and under the terms of this License, through a +publicly available network server or other readily accessible means, +then you must either (1) cause the Corresponding Source to be so +available, or (2) arrange to deprive yourself of the benefit of the +patent license for this particular work, or (3) arrange, in a manner +consistent with the requirements of this License, to extend the patent +license to downstream recipients. "Knowingly relying" means you have +actual knowledge that, but for the patent license, your conveying the +covered work in a country, or your recipient's use of the covered work +in a country, would infringe one or more identifiable patents in that +country that you have reason to believe are valid. + + If, pursuant to or in connection with a single transaction or +arrangement, you convey, or propagate by procuring conveyance of, a +covered work, and grant a patent license to some of the parties +receiving the covered work authorizing them to use, propagate, modify +or convey a specific copy of the covered work, then the patent license +you grant is automatically extended to all recipients of the covered +work and works based on it. + + A patent license is "discriminatory" if it does not include within +the scope of its coverage, prohibits the exercise of, or is +conditioned on the non-exercise of one or more of the rights that are +specifically granted under this License. You may not convey a covered +work if you are a party to an arrangement with a third party that is +in the business of distributing software, under which you make payment +to the third party based on the extent of your activity of conveying +the work, and under which the third party grants, to any of the +parties who would receive the covered work from you, a discriminatory +patent license (a) in connection with copies of the covered work +conveyed by you (or copies made from those copies), or (b) primarily +for and in connection with specific products or compilations that +contain the covered work, unless you entered into that arrangement, +or that patent license was granted, prior to 28 March 2007. + + Nothing in this License shall be construed as excluding or limiting +any implied license or other defenses to infringement that may +otherwise be available to you under applicable patent law. + + 12. No Surrender of Others' Freedom. + + If conditions are imposed on you (whether by court order, agreement or +otherwise) that contradict the conditions of this License, they do not +excuse you from the conditions of this License. If you cannot convey a +covered work so as to satisfy simultaneously your obligations under this +License and any other pertinent obligations, then as a consequence you may +not convey it at all. For example, if you agree to terms that obligate you +to collect a royalty for further conveying from those to whom you convey +the Program, the only way you could satisfy both those terms and this +License would be to refrain entirely from conveying the Program. + + 13. Use with the GNU Affero General Public License. + + Notwithstanding any other provision of this License, you have +permission to link or combine any covered work with a work licensed +under version 3 of the GNU Affero General Public License into a single +combined work, and to convey the resulting work. The terms of this +License will continue to apply to the part which is the covered work, +but the special requirements of the GNU Affero General Public License, +section 13, concerning interaction through a network will apply to the +combination as such. + + 14. Revised Versions of this License. + + The Free Software Foundation may publish revised and/or new versions of +the GNU General Public License from time to time. Such new versions will +be similar in spirit to the present version, but may differ in detail to +address new problems or concerns. + + Each version is given a distinguishing version number. If the +Program specifies that a certain numbered version of the GNU General +Public License "or any later version" applies to it, you have the +option of following the terms and conditions either of that numbered +version or of any later version published by the Free Software +Foundation. If the Program does not specify a version number of the +GNU General Public License, you may choose any version ever published +by the Free Software Foundation. + + If the Program specifies that a proxy can decide which future +versions of the GNU General Public License can be used, that proxy's +public statement of acceptance of a version permanently authorizes you +to choose that version for the Program. + + Later license versions may give you additional or different +permissions. However, no additional obligations are imposed on any +author or copyright holder as a result of your choosing to follow a +later version. + + 15. Disclaimer of Warranty. + + THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY +APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT +HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY +OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, +THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR +PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM +IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF +ALL NECESSARY SERVICING, REPAIR OR CORRECTION. + + 16. Limitation of Liability. + + IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING +WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS +THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY +GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE +USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF +DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD +PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), +EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF +SUCH DAMAGES. + + 17. Interpretation of Sections 15 and 16. + + If the disclaimer of warranty and limitation of liability provided +above cannot be given local legal effect according to their terms, +reviewing courts shall apply local law that most closely approximates +an absolute waiver of all civil liability in connection with the +Program, unless a warranty or assumption of liability accompanies a +copy of the Program in return for a fee. + + END OF TERMS AND CONDITIONS + + How to Apply These Terms to Your New Programs + + If you develop a new program, and you want it to be of the greatest +possible use to the public, the best way to achieve this is to make it +free software which everyone can redistribute and change under these terms. + + To do so, attach the following notices to the program. It is safest +to attach them to the start of each source file to most effectively +state the exclusion of warranty; and each file should have at least +the "copyright" line and a pointer to where the full notice is found. + + + Copyright (C) + + This program is free software: you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation, either version 3 of the License, or + (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program. If not, see . + +Also add information on how to contact you by electronic and paper mail. + + If the program does terminal interaction, make it output a short +notice like this when it starts in an interactive mode: + + Copyright (C) + This program comes with ABSOLUTELY NO WARRANTY; for details type `show w'. + This is free software, and you are welcome to redistribute it + under certain conditions; type `show c' for details. + +The hypothetical commands `show w' and `show c' should show the appropriate +parts of the General Public License. Of course, your program's commands +might be different; for a GUI interface, you would use an "about box". + + You should also get your employer (if you work as a programmer) or school, +if any, to sign a "copyright disclaimer" for the program, if necessary. +For more information on this, and how to apply and follow the GNU GPL, see +. + + The GNU General Public License does not permit incorporating your program +into proprietary programs. If your program is a subroutine library, you +may consider it more useful to permit linking proprietary applications with +the library. If this is what you want to do, use the GNU Lesser General +Public License instead of this License. But first, please read +. diff --git a/ComfyUI-Impact-Pack/README.md b/ComfyUI-Impact-Pack/README.md new file mode 100644 index 0000000000000000000000000000000000000000..d162cf756f9b0b39e3536e940ff81e87b63d7fa9 --- /dev/null +++ b/ComfyUI-Impact-Pack/README.md @@ -0,0 +1,506 @@ +[![Youtube Badge](https://img.shields.io/badge/Youtube-FF0000?style=for-the-badge&logo=Youtube&logoColor=white&link=https://www.youtube.com/watch?v=AccoxDZIg3Y&list=PL_Ej2RDzjQLGfEeizq4GISeY3FtVyFmGP)](https://www.youtube.com/watch?v=AccoxDZIg3Y&list=PL_Ej2RDzjQLGfEeizq4GISeY3FtVyFmGP) + +# ComfyUI-Impact-Pack + +**Custom nodes pack for ComfyUI** +This custom node helps to conveniently enhance images through Detector, Detailer, Upscaler, Pipe, and more. + + +## NOTICE +* V7.6: Automatic installation is no longer supported. Please install using ComfyUI-Manager, or manually install requirements.txt and run install.py to complete the installation. +* V7.0: Supports Switch based on Execution Model Inversion. +* V6.0: Supports FLUX.1 model in Impact KSampler, Detailers, PreviewBridgeLatent +* V5.0: It is no longer compatible with versions of ComfyUI before 2024.04.08. +* V4.87.4: Update to a version of ComfyUI after 2024.04.08 for proper functionality. +* V4.85: Incompatible with the outdated **ComfyUI IPAdapter Plus**. (A version dated March 24th or later is required.) +* V4.77: Compatibility patch applied. Requires ComfyUI version (Oct. 8th) or later. +* V4.73.3: ControlNetApply (SEGS) supports AnimateDiff +* V4.20.1: Due to the feature update in `RegionalSampler`, the parameter order has changed, causing malfunctions in previously created `RegionalSamplers`. Please adjust the parameters accordingly. +* V4.12: `MASKS` is changed to `MASK`. +* V4.7.2 isn't compatible with old version of `ControlNet Auxiliary Preprocessor`. If you will use `MediaPipe FaceMesh to SEGS` update to latest version(Sep. 17th). +* Selection weight syntax is changed(: -> ::) since V3.16. ([tutorial](https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/ImpactWildcardProcessor.md)) +* Starting from V3.6, requires latest version(Aug 8, 9ccc965) of ComfyUI. +* **In versions below V3.3.1, there was an issue with the image quality generated after using the UltralyticsDetectorProvider. Please make sure to upgrade to a newer version.** +* Starting from V3.0, nodes related to `mmdet` are optional nodes that are activated only based on the configuration settings. + - Through ComfyUI-Impact-Subpack, you can utilize UltralyticsDetectorProvider to access various detection models. +* Between versions 2.22 and 2.21, there is partial compatibility loss regarding the Detailer workflow. If you continue to use the existing workflow, errors may occur during execution. An additional output called "enhanced_alpha_list" has been added to Detailer-related nodes. +* The permission error related to cv2 that occurred during the installation of Impact Pack has been patched in version 2.21.4. However, please note that the latest versions of ComfyUI and ComfyUI-Manager are required. +* The "PreviewBridge" feature may not function correctly on ComfyUI versions released before July 1, 2023. +* Attempting to load the "ComfyUI-Impact-Pack" on ComfyUI versions released before June 27, 2023, will result in a failure. +* With the addition of wildcard support in FaceDetailer, the structure of DETAILER_PIPE-related nodes and Detailer nodes has changed. There may be malfunctions when using the existing workflow. + + +## Custom Nodes +### [Detector nodes](https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/detectors.md) + * `SAMLoader` - Loads the SAM model. + * `UltralyticsDetectorProvider` - Loads the Ultralystics model to provide SEGM_DETECTOR, BBOX_DETECTOR. + - Unlike `MMDetDetectorProvider`, for segm models, `BBOX_DETECTOR` is also provided. + - The various models available in UltralyticsDetectorProvider can be downloaded through **ComfyUI-Manager**. + * `ONNXDetectorProvider` - Loads the ONNX model to provide BBOX_DETECTOR. + * `CLIPSegDetectorProvider` - Wrapper for CLIPSeg to provide BBOX_DETECTOR. + * You need to install the ComfyUI-CLIPSeg node extension. + * `SEGM Detector (combined)` - Detects segmentation and returns a mask from the input image. + * `BBOX Detector (combined)` - Detects bounding boxes and returns a mask from the input image. + * `SAMDetector (combined)` - Utilizes the SAM technology to extract the segment at the location indicated by the input SEGS on the input image and outputs it as a unified mask. + * `SAMDetector (Segmented)` - It is similar to `SAMDetector (combined)`, but it separates and outputs the detected segments. Multiple segments can be found for the same detected area, and currently, a policy is in place to group them arbitrarily in sets of three. This aspect is expected to be improved in the future. + * As a result, it outputs the `combined_mask`, which is a unified mask, and `batch_masks`, which are multiple masks grouped together in batch form. + * While `batch_masks` may not be completely separated, it provides functionality to perform some level of segmentation. + * `Simple Detector (SEGS)` - Operating primarily with `BBOX_DETECTOR`, and with the additional provision of `SAM_MODEL` or `SEGM_DETECTOR`, this node internally generates improved SEGS through mask operations on both *bbox* and *silhouette*. It serves as a convenient tool to simplify a somewhat intricate workflow. + +### ControlNet, IPAdapter + * `ControlNetApply (SEGS)` - To apply ControlNet in SEGS, you need to use the Preprocessor Provider node from the Inspire Pack to utilize this node. + * `segs_preprocessor` and `control_image` can be selectively applied. If an `control_image` is given, `segs_preprocessor` will be ignored. + * If set to `control_image`, you can preview the cropped cnet image through `SEGSPreview (CNET Image)`. Images generated by `segs_preprocessor` should be verified through the `cnet_images` output of each Detailer. + * The `segs_preprocessor` operates by applying preprocessing on-the-fly based on the cropped image during the detailing process, while `control_image` will be cropped and used as input to `ControlNetApply (SEGS)`. + * `ControlNetClear (SEGS)` - Clear applied ControlNet in SEGS + * `IPAdapterApply (SEGS)` - To apply IPAdapter in SEGS, you need to use the Preprocessor Provider node from the Inspire Pack to utilize this node. + +### Mask operation + * `Pixelwise(SEGS & SEGS)` - Performs a 'pixelwise and' operation between two SEGS. + * `Pixelwise(SEGS - SEGS)` - Subtracts one SEGS from another. + * `Pixelwise(SEGS & MASK)` - Performs a pixelwise AND operation between SEGS and MASK. + * `Pixelwise(SEGS & MASKS ForEach)` - Performs a pixelwise AND operation between SEGS and MASKS. + * Please note that this operation is performed with batches of MASKS, not just a single MASK. + * `Pixelwise(MASK & MASK)` - Performs a 'pixelwise and' operation between two masks. + * `Pixelwise(MASK - MASK)` - Subtracts one mask from another. + * `Pixelwise(MASK + MASK)` - Combine two masks. + * `SEGM Detector (SEGS)` - Detects segmentation and returns SEGS from the input image. + * `BBOX Detector (SEGS)` - Detects bounding boxes and returns SEGS from the input image. + * `Dilate Mask` - Dilate Mask. + * Support erosion for negative value. + * `Gaussian Blur Mask` - Apply Gaussian Blur to Mask. You can utilize this for mask feathering. + +### [Detailer nodes](https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/detailers.md) + * `Detailer (SEGS)` - Refines the image based on SEGS. + * `DetailerDebug (SEGS)` - Refines the image based on SEGS. Additionally, it provides the ability to monitor the cropped image and the refined image of the cropped image. + * To prevent regeneration caused by the seed that does not change every time when using 'external_seed', please disable the 'seed random generate' option in the 'Detailer...' node. + * `MASK to SEGS` - Generates SEGS based on the mask. + * `MASK to SEGS For AnimateDiff` - Generates SEGS based on the mask for AnimateDiff. + * When using a single mask, convert it to SEGS to apply it to the entire frame. + * When using a batch mask, the contour fill feature is disabled. + * `MediaPipe FaceMesh to SEGS` - Separate each landmark from the mediapipe facemesh image to create labeled SEGS. + * Usually, the size of images created through the MediaPipe facemesh preprocessor is downscaled. It resizes the MediaPipe facemesh image to the original size given as reference_image_opt for matching sizes during processing. + * `ToBinaryMask` - Separates the mask generated with alpha values between 0 and 255 into 0 and 255. The non-zero parts are always set to 255. + * `Masks to Mask List` - This node converts the MASKS in batch form to a list of individual masks. + * `Mask List to Masks` - This node converts the MASK list to MASK batch form. + * `EmptySEGS` - Provides an empty SEGS. + * `MaskPainter` - Provides a feature to draw masks. + * `FaceDetailer` - Easily detects faces and improves them. + * `FaceDetailer (pipe)` - Easily detects faces and improves them (for multipass). + * `MaskDetailer (pipe)` - This is a simple inpaint node that applies the Detailer to the mask area. + + * `FromDetailer (SDXL/pipe)`, `BasicPipe -> DetailerPipe (SDXL)`, `Edit DetailerPipe (SDXL)` - These are pipe functions used in Detailer for utilizing the refiner model of SDXL. + * `Any PIPE -> BasicPipe` - Convert the PIPE Value of other custom nodes that are not BASIC_PIPE but internally have the same structure as BASIC_PIPE to BASIC_PIPE. If an incompatible type is applied, it may cause runtime errors. + +### SEGS Manipulation nodes + * `SEGSDetailer` - Performs detailed work on SEGS without pasting it back onto the original image. + * `SEGSPaste` - Pastes the results of SEGS onto the original image. + * If `ref_image_opt` is present, the images contained within SEGS are ignored. Instead, the image within `ref_image_opt` corresponding to the crop area of SEGS is taken and pasted. The size of the image in `ref_image_opt` should be the same as the original image size. + * This node can be used in conjunction with the processing results of AnimateDiff. + * `SEGSPreview` - Provides a preview of SEGS. + * This option is used to preview the improved image through `SEGSDetailer` before merging it into the original. Prior to going through ```SEGSDetailer```, SEGS only contains mask information without image information. If fallback_image_opt is connected to the original image, SEGS without image information will generate a preview using the original image. However, if SEGS already contains image information, fallback_image_opt will be ignored. + * This node can be used in conjunction with the processing results of AnimateDiff. + * `SEGSPreview (CNET Image)` - Show images configured with `ControlNetApply (SEGS)` for debugging purposes. + * `SEGSToImageList` - Convert SEGS To Image List + * `SEGSToMaskList` - Convert SEGS To Mask List + * `SEGS Filter (label)` - This node filters SEGS based on the label of the detected areas. + * `SEGS Filter (ordered)` - This node sorts SEGS based on size and position and retrieves SEGs within a certain range. + * `SEGS Filter (range)` - This node retrieves only SEGs from SEGS that have a size and position within a certain range. + * `SEGS Assign (label)` - Assign labels sequentially to SEGS. This node is useful when used with `[LAB]` of FaceDetailer. + * `SEGSConcat` - Concatenate segs1 and segs2. If source shape of segs1 and segs2 are different from segs2 will be ignored. + * `SEGS Merge` - SEGS contains multiple SEGs. SEGS Merge integrates several SEGs into a single merged SEG. The label is changed to `merged` and the confidence becomes the minimum confidence. The applied controlnet and cropped_image are removed. + * `Picker (SEGS)` - Among the input SEGS, you can select a specific SEG through a dialog. If no SEG is selected, it outputs an empty SEGS. Increasing the batch_size of SEGSDetailer can be used for the purpose of selecting from the candidates. + * `Set Default Image For SEGS` - Set a default image for SEGS. SEGS with images set this way do not need to have a fallback image set. When override is set to false, the original image is preserved. + * `Remove Image from SEGS` - Remove the image set for the SEGS that has been configured by "Set Default Image for SEGS" or SEGSDetailer. When the image for the SEGS is removed, the Detailer node will operate based on the currently processed image instead of the SEGS. + * `Make Tile SEGS` - [experimental] Create SEGS in the form of tiles from an image to facilitate experiments for Tiled Upscale using the Detailer. + * The `filter_in_segs_opt` and `filter_out_segs_opt` are optional inputs. If these inputs are provided, when creating the tiles, the mask for each tile is generated by overlapping with the mask of `filter_in_segs_opt` and excluding the overlap with the mask of `filter_out_segs_opt`. Tiles with an empty mask will not be created as SEGS. + * `Dilate Mask (SEGS)` - Dilate/Erosion Mask in SEGS + * `Gaussian Blur Mask (SEGS)` - Apply Gaussian Blur to Mask in SEGS + * `SEGS_ELT Manipulation` - experimental nodes + * `DecomposeSEGS` - Decompose SEGS to allow for detailed manipulation. + * `AssembleSEGS` - Reassemble the decomposed SEGS. + * `From SEG_ELT` - Extract detailed information from SEG_ELT. + * `Edit SEG_ELT` - Modify some of the information in SEG_ELT. + * `Dilate SEG_ELT` - Dilate the mask of SEG_ELT. + * `From SEG_ELT` bbox - Extract coordinate from bbox in SEG_ELT + * `From SEG_ELT` crop_region - Extract coordinate from crop_region in SEG_ELT + * `Count Elt in SEGS` - Number of Elts ins SEGS + +### Pipe nodes + * `ToDetailerPipe`, `FromDetailerPipe` - These nodes are used to bundle multiple inputs used in the detailer, such as models and vae, ..., into a single DETAILER_PIPE or extract the elements that are bundled in the DETAILER_PIPE. + * `ToBasicPipe`, `FromBasicPipe` - These nodes are used to bundle model, clip, vae, positive conditioning, and negative conditioning into a single BASIC_PIPE, or extract each element from the BASIC_PIPE. + * `EditBasicPipe`, `EditDetailerPipe` - These nodes are used to replace some elements in BASIC_PIPE or DETAILER_PIPE. + * `FromDetailerPipe_v2`, `FromBasicPipe_v2` - It has the same functionality as `FromDetailerPipe` and `FromBasicPipe`, but it has an additional output that directly exports the input pipe. It is useful when editing EditBasicPipe and EditDetailerPipe. +* `Latent Scale (on Pixel Space)` - This node converts latent to pixel space, upscales it, and then converts it back to latent. + * If upscale_model_opt is provided, it uses the model to upscale the pixel and then downscales it using the interpolation method provided in scale_method to the target resolution. +* `PixelKSampleUpscalerProvider` - An upscaler is provided that converts latent to pixels using VAEDecode, performs upscaling, converts back to latent using VAEEncode, and then performs k-sampling. This upscaler can be attached to nodes such as `Iterative Upscale` for use. + * Similar to `Latent Scale (on Pixel Space)`, if upscale_model_opt is provided, it performs pixel upscaling using the model. +* `PixelTiledKSampleUpscalerProvider` - It is similar to `PixelKSampleUpscalerProvider`, but it uses `ComfyUI_TiledKSampler` and Tiled VAE Decoder/Encoder to avoid GPU VRAM issues at high resolutions. + * You need to install the [BlenderNeko/ComfyUI_TiledKSampler](https://github.com/BlenderNeko/ComfyUI_TiledKSampler) node extension. + +### PK_HOOK + * `DenoiseScheduleHookProvider` - IterativeUpscale provides a hook that gradually changes the denoise to target_denoise as the iterative-step progresses. + * `CfgScheduleHookProvider` - IterativeUpscale provides a hook that gradually changes the cfg to target_cfg as the iterative-step progresses. + * `StepsScheduleHookProvider` - IterativeUpscale provides a hook that gradually changes the sampling-steps to target_steps as the iterative-step progresses. + * `NoiseInjectionHookProvider` - During each iteration of IterativeUpscale, noise is injected into the latent space while varying the strength according to a schedule. + * You need to install the [BlenderNeko/ComfyUI_Noise](https://github.com/BlenderNeko/ComfyUI_Noise) node extension. + * The seed serves as the initial value required for generating noise, and it increments by 1 with each iteration as the process unfolds. + * The source determines the types of CPU noise and GPU noise to be configured. + * Currently, there is only a simple schedule available, where the strength of the noise varies from start_strength to end_strength during the progression of each iteration. + * `UnsamplerHookProvider` - Apply Unsampler during each iteration. To use this node, ComfyUI_Noise must be installed. + * `PixelKSampleHookCombine` - This is used to connect two PK_HOOKs. hook1 is executed first and then hook2 is executed. + * If you want to simultaneously change cfg and denoise, you can combine the PK_HOOKs of CfgScheduleHookProvider and PixelKSampleHookCombine. + +### DETAILER_HOOK + * `NoiseInjectionDetailerHookProvider` - The `detailer_hook` is a hook in the `Detailer` that injects noise during the processing of each SEGS. + * `UnsamplerDetailerHookProvider` - Apply Unsampler during each cycle. To use this node, ComfyUI_Noise must be installed. + * `DenoiseSchedulerDetailerHookProvider` - During the progress of the cycle, the detailer's denoise is altered up to the `target_denoise`. + * `CoreMLDetailerHookProvider` - CoreML supports only 512x512, 512x768, 768x512, 768x768 size sampling. CoreMLDetailerHookProvider precisely fixes the upscale of the crop_region to this size. When using this hook, it will always be selected size, regardless of the guide_size. However, if the guide_size is too small, skipping will occur. + * `DetailerHookCombine` - This is used to connect two DETAILER_HOOKs. Similar to PixelKSampleHookCombine. + * `SEGSOrderedFilterDetailerHook`, SEGSRangeFilterDetailerHook, SEGSLabelFilterDetailerHook - There are a wrapper node that provides SEGSFilter nodes to be applied in FaceDetailer or Detector by creating DETAILER_HOOK. + * `PreviewDetailerHook` - Connecting this hook node helps provide assistance for viewing previews whenever SEGS Detailing tasks are completed. When working with a large number of SEGS, such as Make Tile SEGS, it allows for monitoring the situation as improvements progress incrementally. + * Since this is the hook applied when pasting onto the original image, it has no effect on nodes like `SEGSDetailer`. + * `VariationNoiseDetailerHookProvider` - Apply variation seed to the detailer. It can be applied in multiple stages through combine. + +### Iterative Upscale nodes + * `Iterative Upscale (Latent/on Pixel Space)` - The upscaler takes the input upscaler and splits the scale_factor into steps, then iteratively performs upscaling. + This takes latent as input and outputs latent as the result. + * `Iterative Upscale (Image)` - The upscaler takes the input upscaler and splits the scale_factor into steps, then iteratively performs upscaling. This takes image as input and outputs image as the result. + * Internally, this node uses 'Iterative Upscale (Latent)'. + +### TwoSamplers nodes +* `TwoSamplersForMask` - This node can apply two samplers depending on the mask area. The base_sampler is applied to the area where the mask is 0, while the mask_sampler is applied to the area where the mask is 1. + * Note: The latent encoded through VAEEncodeForInpaint cannot be used. +* `KSamplerProvider` - This is a wrapper that enables KSampler to be used in TwoSamplersForMask TwoSamplersForMaskUpscalerProvider. +* `TiledKSamplerProvider` - ComfyUI_TiledKSampler is a wrapper that provides KSAMPLER. + * You need to install the [BlenderNeko/ComfyUI_TiledKSampler](https://github.com/BlenderNeko/ComfyUI_TiledKSampler) node extension. + +* `TwoAdvancedSamplersForMask` - TwoSamplersForMask is similar to TwoAdvancedSamplersForMask, but they differ in their operation. TwoSamplersForMask performs sampling in the mask area only after all the samples in the base area are finished. On the other hand, TwoAdvancedSamplersForMask performs sampling in both the base area and the mask area sequentially at each step. +* `KSamplerAdvancedProvider` - This is a wrapper that enables KSampler to be used in TwoAdvancedSamplersForMask, RegionalSampler. + * sigma_factor: By multiplying the denoise schedule by the sigma_factor, you can adjust the amount of denoising based on the configured denoise. + +* `TwoSamplersForMaskUpscalerProvider` - This is an Upscaler that extends TwoSamplersForMask to be used in Iterative Upscale. + * TwoSamplersForMaskUpscalerProviderPipe - pipe version of TwoSamplersForMaskUpscalerProvider. + +### Image Utils + * `PreviewBridge (image)` - This custom node can be used with a bridge for image when using the MaskEditor feature of Clipspace. + * `PreviewBridge (latent)` - This custom node can be used with a bridge for latent image when using the MaskEditor feature of Clipspace. + * If a latent with a mask is provided as input, it displays the mask. Additionally, the mask output provides the mask set in the latent. + * If a latent without a mask is provided as input, it outputs the original latent as is, but the mask output provides an output with the entire region set as a mask. + * When set mask through MaskEditor, a mask is applied to the latent, and the output includes the stored mask. The same mask is also output as the mask output. + * When connected to `vae_opt`, it takes higher priority than the `preview_method`. + * `ImageSender`, `ImageReceiver` - The images generated in ImageSender are automatically sent to the ImageReceiver with the same link_id. + * `LatentSender`, `LatentReceiver` - The latent generated in LatentSender are automatically sent to the LatentReceiver with the same link_id. + * Furthermore, LatentSender is implemented with PreviewLatent, which stores the latent in payload form within the image thumbnail. + * Due to the current structure of ComfyUI, it is unable to distinguish between SDXL latent and SD1.5/SD2.1 latent. Therefore, it generates thumbnails by decoding them using the SD1.5 method. + +### Switch nodes + * `Switch (image,mask)`, `Switch (latent)`, `Switch (SEGS)` - Among multiple inputs, it selects the input designated by the selector and outputs it. The first input must be provided, while the others are optional. However, if the input specified by the selector is not connected, an error may occur. + * `Switch (Any)` - This is a Switch node that takes an arbitrary number of inputs and produces a single output. Its type is determined when connected to any node, and connecting inputs increases the available slots for connections. + * `Inversed Switch (Any)` - In contrast to `Switch (Any)`, it takes a single input and outputs one of many. + * NOTE: See this [tutorial](https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/switch.md) + +### [Wildcards](http://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/ImpactWildcard.md) nodes + * These are nodes that supports syntax in the form of `__wildcard-name__` and dynamic prompt syntax like `{a|b|c}`. + * Wildcard files can be used by placing `.txt` or `.yaml` files under either `ComfyUI-Impact-Pack/wildcards` or `ComfyUI-Impact-Pack/custom_wildcards` paths. + * You can download and use [Wildcard YAML](https://civitai.com/models/138970/billions-of-wildcards-all-in-one) files in this format. + * After the first execution, you can change the custom wildcards path in the `custom_wildcards` entry within the `ComfyUI-Impact-Pack/impact-pack.ini` file created. + * `ImpactWildcardProcessor` - The text is generated by processing the wildcard in the Text. If the mode is set to "populate", a dynamic prompt is generated with each execution and the input is filled in the second textbox. If the mode is set to "fixed", the content of the second textbox remains unchanged. + * When an image is generated with the "fixed" mode, the prompt used for that particular generation is stored in the metadata. + * `ImpactWildcardEncode` - Similar to ImpactWildcardProcessor, this provides the loading functionality of LoRAs (e.g. ``). Populated prompts are encoded using the clip after all the lora loading is done. + * If the `Inspire Pack` is installed, you can use **Lora Block Weight** in the form of `LBW=lbw spec;` + * ``, ``, `` + +### Regional Sampling + * These nodes offer the capability to divide regions and perform partial sampling using a mask. Unlike TwoSamplersForMask, sampling for each region is applied during each step. + * `RegionalPrompt` - This node combines a **mask** for specifying regions and the **sampler** to apply to each region to create `REGIONAL_PROMPTS`. + * `CombineRegionalPrompts` - Combine multiple `REGIONAL_PROMPTS` to create a single `REGIONAL_PROMPTS`. + * `RegionalSampler` - This node performs sampling using a base sampler and regional prompts. Sampling by the base sampler is executed at each step, while sampling for each region is performed through the sampler bound to each region. + * overlap_factor - Specifies the amount of overlap for each region to blend well with the area outside the mask. + * restore_latent - When sampling each region, restore the areas outside the mask to the base latent, preventing additional noise from being introduced outside the mask during region sampling. + * `RegionalSamplerAdvanced` - This is the Advanced version of the RegionalSampler. You can control it using `step` instead of `denoise`. + > NOTE: The `sde` sampler and `uni_pc` sampler introduce additional noise during each step of the sampling process. To mitigate this, when sampling each region, the `uni_pc` sampler applies additional `dpmpp_fast`, and the sde sampler applies the `dpmpp_2m` sampler as an additional measure. + + +### Impact KSampler + * These samplers support basic_pipe and AYS scheduler + * `KSampler (pipe)` - pipe version of KSampler + * `KSampler (advanced/pipe)` - pipe version of KSamplerAdvacned + * When converting the scheduler widget to input, refer to the `Impact Scheduler Adapter` node to resolve compatibility issues. + * `GITSScheduler Func Provider` - provider scheduler function for GITSScheduler + + +### Batch/List Util + * `Image Batch to Image List` - Convert Image batch to Image List + - You can use images generated in a multi batch to handle them + * `Image List to Image Batch` - Convert Image List to Image Batch + * `Make Image List` - Convert multiple images into a single image list + * `Make Image Batch` - Convert multiple images into a single image batch + - The input of images can be scaled up as needed + * `Masks to Mask List`, `Mask List to Masks`, `Make Mask List`, `Make Mask Batch` - It has the same functionality as the nodes above, but uses mask as input instead of image. + * `Flatten Mask Batch` - Flattens a Mask Batch into a single Mask. Normal operation is not guaranteed for non-binary masks. + * `Make List (Any)` - Create a list with arbitrary values. + +### Logics (experimental) + * These nodes are experimental nodes designed to implement the logic for loops and dynamic switching. + * `ImpactCompare`, `ImpactConditionalBranch`, `ImpactConditionalBranchSelMode`, `ImpactInt`, `ImpactBoolean`, `ImpactValueSender`, `ImpactValueReceiver`, `ImpactImageInfo`, `ImpactMinMax`, `ImpactNeg`, `ImpactConditionalStopIteration` + * `ImpactIsNotEmptySEGS` - This node returns `true` only if the input SEGS is not empty. + * `ImpactIfNone` - Returns `true` if any_input is None, and returns `false` if it is not None. + * `Queue Trigger` - When this node is executed, it adds a new queue to assist with repetitive tasks. It will only execute if the signal's status changes. + * `Queue Trigger (Countdown)` - Like the Queue Trigger, it adds a queue, but only adds it if it's greater than 1, and decrements the count by one each time it runs. + * `Sleep` - Waits for the specified time (in seconds). + * `Set Widget Value` - This node sets one of the optional inputs to the specified node's widget. An error may occur if the types do not match. + * `Set Mute State` - This node changes the mute state of a specific node. + * `Control Bridge` - This node modifies the state of the connected control nodes based on the `mode` and `behavior` . If there are nodes that require a change, the current execution is paused, the mute status is updated, and a new prompt queue is inserted. + * When the `mode` is `active`, it makes the connected control nodes active regardless of the behavior. + * When the `mode` is `Bypass/Mute`, it changes the state of the connected nodes based on whether the behavior is `Bypass` or `Mute`. + * **Limitation**: Due to these characteristics, it does not function correctly when the batch count exceeds 1. Additionally, it does not guarantee proper operation when the seed is randomized or when the state of nodes is altered by actions such as `Queue Trigger`, `Set Widget Value`, `Set Mute`, before the Control Bridge. + * When utilizing this node, please structure the workflow in such a way that `Queue Trigger`, `Set Widget Value`, `Set Mute State`, and similar actions are executed at the end of the workflow. + * If you want to change the value of the seed at each iteration, please ensure that Set Widget Value is executed at the end of the workflow instead of using randomization. + * It is not a problem if the seed changes due to randomization as long as it occurs after the Control Bridge section. + * `Remote Boolean (on prompt)`, `Remote Int (on prompt)` - At the start of the prompt, this node forcibly sets the `widget_value` of `node_id`. It is disregarded if the target widget type is different. + * You can find the `node_id` by checking through [ComfyUI-Manager](https://github.com/ltdrdata/ComfyUI-Manager) using the format `Badge: #ID Nickname`. + * Experimental set of nodes for implementing loop functionality (tutorial to be prepared later / [example workflow](test/loop-test.json)). + +### HuggingFace nodes + * These nodes provide functionalities based on HuggingFace repository models. + * The path where the HuggingFace model cache is stored can be changed through the `HF_HOME` environment variable. + * `HF Transformers Classifier Provider` - This is a node that provides a classifier based on HuggingFace's transformers models. + * The 'repo id' parameter should contain HuggingFace's repo id. When `preset_repo_id` is set to `Manual repo id`, use the manually entered repo id in `manual_repo_id`. + * e.g. 'rizvandwiki/gender-classification-2' is a repository that provides a model for gender classification. + * `SEGS Classify` - This node utilizes the `TRANSFORMERS_CLASSIFIER` loaded with 'HF Transformers Classifier Provider' to classify `SEGS`. + * The 'expr' allows for forms like `label > number`, and in the case of `preset_expr` being `Manual expr`, it uses the expression entered in `manual_expr`. + * For example, in the case of `male <= 0.4`, if the score of the `male` label in the classification result is less than or equal to 0.4, it is categorized as `filtered_SEGS`, otherwise, it is categorized as `remained_SEGS`. + * For supported labels, please refer to the `config.json` of the respective HuggingFace repository. + * `#Female` and `#Male` are symbols that group multiple labels such as `Female, women, woman, ...`, for convenience, rather than being single labels. + +### Etc nodes + * `Impact Scheduler Adapter` - With the addition of AYS to the scheduler of the Impact Pack and Inspire Pack, there is an issue of incompatibility when the existing scheduler widget is converted to input. The Impact Scheduler Adapter allows for an indirect connection to be possible. + * `StringListToString` - Convert String List to String + * `WildcardPromptFromString` - Create labeled wildcard for detailer from string. + * This node works well when used with MakeTileSEGS. [[Link](https://github.com/ltdrdata/ComfyUI-Impact-Pack/pull/536#discussion_r1586060779)] + + * `String Selector` - It selects and returns a portion of the string. When `multiline` mode is disabled, it simply returns the string of the line pointed to by the selector. When `multiline` mode is enabled, it divides the string based on lines that start with `#` and returns them. If the `select` value is larger than the number of items, it will start counting from the first line again and return accordingly. + * `Combine Conditionings` - It takes multiple conditionings as input and combines them into a single conditioning. + * `Concat Conditionings` - It takes multiple conditionings as input and concat them into a single conditioning. + * `Negative Cond Placeholder` - Models like FLUX.1 do not use Negative Conditioning. This is a placeholder node for them. You can use FLUX.1 by replacing the Negative Conditioning used in Impact KSampler, KSampler (Inspire), and Detailer with this node. + * `Execution Order Controller` - A helper node that can forcibly control the execution order of nodes. + * Connect the output of the node that should be executed first to the signal, and make the input of the node that should be executed later pass through this node. + + +## MMDet nodes (DEPRECATED) - Don't use these nodes +* MMDetDetectorProvider - Loads the MMDet model to provide BBOX_DETECTOR and SEGM_DETECTOR. +* To use the existing MMDetDetectorProvider, you need to enable the MMDet usage configuration. + + +## Feature +* `Interactive SAM Detector (Clipspace)` - When you right-click on a node that has 'MASK' and 'IMAGE' outputs, a context menu will open. From this menu, you can either open a dialog to create a SAM Mask using 'Open in SAM Detector', or copy the content (likely mask data) using 'Copy (Clipspace)' and generate a mask using 'Impact SAM Detector' from the clipspace menu, and then paste it using 'Paste (Clipspace)'. +* Providing a feature to detect errors that occur when mixing models and clips from checkpoints such as `SDXL Base`, `SDXL Refiner`, `SD1.x`, `SD2.x` during sample execution, and reporting appropriate errors. + + +## Deprecated +* The following nodes have been kept only for compatibility with existing workflows, and are no longer supported. Please replace them with new nodes. + * ONNX Detector (SEGS) - BBOX Detector (SEGS) + * MMDetLoader -> MMDetDetectorProvider + * SegsMaskCombine -> SEGS to MASK (combined) + * BboxDetectorForEach -> BBOX Detector (SEGS) + * SegmDetectorForEach -> SEGM Detector (SEGS) + * BboxDetectorCombined -> BBOX Detector (combined) + * SegmDetectorCombined -> SEGM Detector (combined) + * MaskPainter -> PreviewBridge +* To use the existing deprecated legacy nodes, you need to enable the MMDet usage configuration. + + +## Ultralytics models +* When using ultralytics models, save them separately in `models/ultralytics/bbox` and `models/ultralytics/segm` depending on the type of model. Many models can be downloaded by searching for `ultralytics` in the Model Manager of ComfyUI-Manager. +* huggingface.co/Bingsu/[adetailer](https://huggingface.co/Bingsu/adetailer/tree/main) - You can download face, people detection models, and clothing detection models. +* ultralytics/[assets](https://github.com/ultralytics/assets/releases/) - You can download various types of detection models other than faces or people. +* civitai/[adetailer](https://civitai.com/search/models?sortBy=models_v5&query=adetailer) - You can download various types detection models....Many models are associated with NSFW content. + +## How to activate 'MMDet usage' (DEPRECATED) +* Upon the initial execution, an `impact-pack.ini` file will be generated in the custom_nodes/ComfyUI-Impact-Pack directory. +``` +[default] +dependency_version = 2 +mmdet_skip = True +``` +* Change `mmdet_skip = True` to `mmdet_skip = False` +``` +[default] +dependency_version = 2 +mmdet_skip = False +``` +* Restart ComfyUI + + +## Installation + +### Install via ComfyUI-Manager (Recommended) +* Search `ComfyUI Impact Pack` in ComfyUI-Manager and click `Install` button. + +### Manual Install (Not Recommended) +1. `cd custom_nodes` +2. `git clone https://github.com/ltdrdata/ComfyUI-Impact-Pack.git` +3. `cd ComfyUI-Impact-Pack` +4. (optional) `git clone https://github.com/ltdrdata/ComfyUI-Impact-Subpack impact_subpack` + * Impact Pack will automatically download subpack during its initial launch. +5. (optional) `python install-manual.py` + * Impact Pack will automatically install its dependencies during its initial launch. + * For the portable version, you should execute the command `..\..\..\python_embeded\python.exe install-manual.py` to run the installation script. +6. Restart ComfyUI + +* NOTE1: If an error occurs during the installation process, please refer to [Troubleshooting Page](troubleshooting/TROUBLESHOOTING.md) for assistance. +* NOTE2: You can use this colab notebook [colab notebook](https://colab.research.google.com/github/ltdrdata/ComfyUI-Impact-Pack/blob/Main/notebook/comfyui_colab_impact_pack.ipynb) to launch it. This notebook automatically downloads the impact pack to the custom_nodes directory, installs the tested dependencies, and runs it. +* NOTE3: If you create an empty file named `skip_download_model` in the `ComfyUI/custom_nodes/` directory, it will skip the model download step during the installation of the impact pack. + + +## Package Dependencies (If you need to manual setup.) + +* pip install + * segment-anything + * ultralytics + * scikit-image + * piexif + * opencv-python + * GitPython + * scipy + * numpy<2 + * dill + * matplotlib + * (optional) onnxruntime + * (deprecated) openmim # for mim + * (deprecated) pycocotools # for mim + +* mim install (deprecated) + * mmcv==2.0.0, mmdet==3.0.0, mmengine==0.7.2 + +* linux packages (ubuntu) + * libgl1-mesa-glx + * libglib2.0-0 + + +## Config example +* Once you run the Impact Pack for the first time, an `impact-pack.ini` file will be automatically generated in the Impact Pack directory. You can modify this configuration file to customize the default behavior. + * `dependency_version` - don't touch this + * `mmdet_skip` - disable MMDet based nodes and legacy nodes if `True` + * `sam_editor_cpu` - use cpu for `SAM editor` instead of gpu + * sam_editor_model: Specify the SAM model for the SAM editor. + * You can download various SAM models using ComfyUI-Manager. + * Path to SAM model: `ComfyUI/models/sams` +``` +[default] +dependency_version = 9 +mmdet_skip = True +sam_editor_cpu = False +sam_editor_model = sam_vit_b_01ec64.pth +``` + + +## Other Materials (auto-download on initial startup) + +* ComfyUI/models/mmdets/bbox <= https://huggingface.co/dustysys/ddetailer/resolve/main/mmdet/bbox/mmdet_anime-face_yolov3.pth +* ComfyUI/models/mmdets/bbox <= https://raw.githubusercontent.com/Bing-su/dddetailer/master/config/mmdet_anime-face_yolov3.py +* ComfyUI/models/sams <= https://dl.fbaipublicfiles.com/segment_anything/sam_vit_b_01ec64.pth + +## Troubleshooting page +* [Troubleshooting Page](troubleshooting/TROUBLESHOOTING.md) + + +## How to use (DDetailer feature) + +#### 1. Basic auto face detection and refine exapmle. +![simple](https://github.com/ltdrdata/ComfyUI-extension-tutorials/raw/Main/ComfyUI-Impact-Pack/images/simple.png) +* The face that has been damaged due to low resolution is restored with high resolution by generating and synthesizing it, in order to restore the details. +* The FaceDetailer node is a combination of a Detector node for face detection and a Detailer node for image enhancement. See the [Advanced Tutorial](https://github.com/ltdrdata/ComfyUI-extension-tutorials/raw/Main/ComfyUI-Impact-Pack/tutorial/advanced.md) for a more detailed explanation. +* Pass the MMDetLoader 's bbox model and the detection model loaded by SAMLoader to FaceDetailer . Since it performs the function of KSampler for image enhancement, it overlaps with KSampler's options. +* The MASK output of FaceDetailer provides a visualization of where the detected and enhanced areas are. + +![simple-orig](https://github.com/ltdrdata/ComfyUI-extension-tutorials/raw/Main/ComfyUI-Impact-Pack/images/simple-original.png) ![simple-refined](https://github.com/ltdrdata/ComfyUI-extension-tutorials/raw/Main/ComfyUI-Impact-Pack/images/simple-refined.png) +* You can see that the face in the image on the left has increased detail as in the image on the right. + +#### 2. 2Pass refine (restore a severely damaged face) +![2pass-workflow-example](https://github.com/ltdrdata/ComfyUI-extension-tutorials/raw/Main/ComfyUI-Impact-Pack/images/2pass-simple.png) +* Although two FaceDetailers can be attached together for a 2-pass configuration, various common inputs used in KSampler can be passed through DETAILER_PIPE, so FaceDetailerPipe can be used to configure easily. +* In 1pass, only rough outline recovery is required, so restore with a reasonable resolution and low options. However, if you increase the dilation at this time, not only the face but also the surrounding parts are included in the recovery range, so it is useful when you need to reshape the face other than the facial part. + +![2pass-example-original](https://github.com/ltdrdata/ComfyUI-extension-tutorials/raw/Main/ComfyUI-Impact-Pack/images/2pass-original.png) ![2pass-example-middle](https://github.com/ltdrdata/ComfyUI-extension-tutorials/raw/Main/ComfyUI-Impact-Pack/images/2pass-1pass.png) ![2pass-example-result](https://github.com/ltdrdata/ComfyUI-extension-tutorials/raw/Main/ComfyUI-Impact-Pack/images/2pass-2pass.png) +* In the first stage, the severely damaged face is restored to some extent, and in the second stage, the details are restored + +#### 3. Face Bbox(bounding box) + Person silhouette segmentation (prevent distortion of the background.) +![combination-workflow-example](https://github.com/ltdrdata/ComfyUI-extension-tutorials/raw/Main/ComfyUI-Impact-Pack/images/combination.jpg) +![combination-example-original](https://github.com/ltdrdata/ComfyUI-extension-tutorials/raw/Main/ComfyUI-Impact-Pack/images/combination-original.png) ![combination-example-refined](https://github.com/ltdrdata/ComfyUI-extension-tutorials/raw/Main/ComfyUI-Impact-Pack/images/combination-refined.png) + +* Facial synthesis that emphasizes details is delicately aligned with the contours of the face, and it can be observed that it does not affect the image outside of the face. + +* The BBoxDetectorForEach node is used to detect faces, and the SAMDetectorCombined node is used to find the segment related to the detected face. By using the Segs & Mask node with the two masks obtained in this way, an accurate mask that intersects based on segs can be generated. If this generated mask is input to the DetailerForEach node, only the target area can be created in high resolution from the image and then composited. + +#### 4. Iterative Upscale +![upscale-workflow-example](https://github.com/ltdrdata/ComfyUI-extension-tutorials/raw/Main/ComfyUI-Impact-Pack/images/upscale-workflow.png) + +* The IterativeUpscale node is a node that enlarges an image/latent by a scale_factor. In this process, the upscale is carried out progressively by dividing it into steps. +* IterativeUpscale takes an Upscaler as an input, similar to a plugin, and uses it during each iteration. PixelKSampleUpscalerProvider is an Upscaler that converts the latent representation to pixel space and applies ksampling. + * The upscale_model_opt is an optional parameter that determines whether to use the upscale function of the model base if available. Using the upscale function of the model base can significantly reduce the number of iterative steps required. If an x2 upscaler is used, the image/latent is first upscaled by a factor of 2 and then downscaled to the target scale at each step before further processing is done. + +* The following image is an image of 304x512 pixels and the same image scaled up to three times its original size using IterativeUpscale. + +![combination-example-original](https://github.com/ltdrdata/ComfyUI-extension-tutorials/raw/Main/ComfyUI-Impact-Pack/images/upscale-original.png) ![combination-example-refined](https://github.com/ltdrdata/ComfyUI-extension-tutorials/raw/Main/ComfyUI-Impact-Pack/images/upscale-3x.png) + + +#### 5. Interactive SAM Detector (Clipspace) + +* When you right-click on the node that outputs 'MASK' and 'IMAGE', a menu called "Open in SAM Detector" appears, as shown in the following picture. Clicking on the menu opens a dialog in SAM's functionality, allowing you to generate a segment mask. +![samdetector-menu](https://github.com/ltdrdata/ComfyUI-extension-tutorials/raw/Main/ComfyUI-Impact-Pack/images/SAMDetector-menu.png) + +* By clicking the left mouse button on a coordinate, a positive prompt in blue color is entered, indicating the area that should be included. Clicking the right mouse button on a coordinate enters a negative prompt in red color, indicating the area that should be excluded. Positive prompts represent the areas that should be included, while negative prompts represent the areas that should be excluded. +* You can remove the points that were added by using the "undo" button. After selecting the points, pressing the "detect" button generates the mask. Additionally, you can adjust the fidelity slider to determine the extent to which the mask belongs to the confidence region. + +![samdetector-dialog](https://github.com/ltdrdata/ComfyUI-extension-tutorials/raw/Main/ComfyUI-Impact-Pack/images/SAMDetector-dialog.jpg) + +* If you opened the dialog through "Open in SAM Detector" from the node, you can directly apply the changes by clicking the "Save to node" button. However, if you opened the dialog through the "clipspace" menu, you can save it to clipspace by clicking the "Save" button. + +![samdetector-result](https://github.com/ltdrdata/ComfyUI-extension-tutorials/raw/Main/ComfyUI-Impact-Pack/images/SAMDetector-result.jpg) + +* When you execute using the reflected mask in the node, you can observe that the image and mask are displayed separately. + + +## Others Tutorials +* [ComfyUI-extension-tutorials/ComfyUI-Impact-Pack](https://github.com/ltdrdata/ComfyUI-extension-tutorials/tree/Main/ComfyUI-Impact-Pack) - You can find various tutorials and workflows on this page. +* [Advanced Tutorial](https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/advanced.md) +* [SAM Application](https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/sam.md) +* [PreviewBridge](https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/previewbridge.md) +* [Mask Pointer](https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/maskpointer.md) +* [ONNX Tutorial](https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/ONNX.md) +* [CLIPSeg Tutorial](https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/clipseg.md) +* [Extreme Highresolution Upscale](https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/extreme-upscale.md) +* [TwoSamplersForMask](https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/TwoSamplers.md) +* [TwoAdvancedSamplersForMask](https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/TwoAdvancedSamplers.md) +* [Advanced Iterative Upscale: PK_HOOK](https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/pk_hook.md) +* [Advanced Iterative Upscale: TwoSamplersForMask Upscale Provider](https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/TwoSamplersUpscale.md) +* [Interactive SAM + PreviewBridge](https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/sam_with_preview_bridge.md) +* [ImageSender/ImageReceiver/LatentSender/LatentReceiver](https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/sender_receiver.md) +* [ImpactWildcardProcessor](https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/ImpactWildcardProcessor.md) + + +## Credits + +ComfyUI/[ComfyUI](https://github.com/comfyanonymous/ComfyUI) - A powerful and modular stable diffusion GUI. + +dustysys/[ddetailer](https://github.com/dustysys/ddetailer) - DDetailer for Stable-diffusion-webUI extension. + +Bing-su/[dddetailer](https://github.com/Bing-su/dddetailer) - The anime-face-detector used in ddetailer has been updated to be compatible with mmdet 3.0.0, and we have also applied a patch to the pycocotools dependency for Windows environment in ddetailer. + +facebook/[segment-anything](https://github.com/facebookresearch/segment-anything) - Segmentation Anything! + +hysts/[anime-face-detector](https://github.com/hysts/anime-face-detector) - Creator of `anime-face_yolov3`, which has impressive performance on a variety of art styles. + +open-mmlab/[mmdetection](https://github.com/open-mmlab/mmdetection) - Object detection toolset. `dd-person_mask2former` was trained via transfer learning using their [R-50 Mask2Former instance segmentation model](https://github.com/open-mmlab/mmdetection/tree/master/configs/mask2former#instance-segmentation) as a base. + +biegert/[ComfyUI-CLIPSeg](https://github.com/biegert/ComfyUI-CLIPSeg) - This is a custom node that enables the use of CLIPSeg technology, which can find segments through prompts, in ComfyUI. + +BlenderNeok/[ComfyUI-TiledKSampler](https://github.com/BlenderNeko/ComfyUI_TiledKSampler) - The tile sampler allows high-resolution sampling even in places with low GPU VRAM. + +BlenderNeok/[ComfyUI_Noise](https://github.com/BlenderNeko/ComfyUI_Noise) - The noise injection feature relies on this function and slerp code for noise variation + +WASasquatch/[was-node-suite-comfyui](https://github.com/WASasquatch/was-node-suite-comfyui) - A powerful custom node extensions of ComfyUI. + +Trung0246/[ComfyUI-0246](https://github.com/Trung0246/ComfyUI-0246) - Nice bypass hack! diff --git a/ComfyUI-Impact-Pack/__init__.py b/ComfyUI-Impact-Pack/__init__.py new file mode 100644 index 0000000000000000000000000000000000000000..2a2203d9982c34234bedcf7a1164d0b04d62cedc --- /dev/null +++ b/ComfyUI-Impact-Pack/__init__.py @@ -0,0 +1,498 @@ +""" +@author: Dr.Lt.Data +@title: Impact Pack +@nickname: Impact Pack +@description: This extension offers various detector nodes and detailer nodes that allow you to configure a workflow that automatically enhances facial details. And provide iterative upscaler. +""" + +import shutil +import folder_paths +import os +import sys +import traceback + +comfy_path = os.path.dirname(folder_paths.__file__) +impact_path = os.path.join(os.path.dirname(__file__)) +subpack_path = os.path.join(os.path.dirname(__file__), "impact_subpack") +modules_path = os.path.join(os.path.dirname(__file__), "modules") + +sys.path.append(modules_path) + +import impact.config +import impact.sample_error_enhancer +print(f"### Loading: ComfyUI-Impact-Pack ({impact.config.version})") + + +sys.path.append(subpack_path) + +# Core +# recheck dependencies for colab +try: + import impact.subpack_nodes # This import must be done before cv2. + + import folder_paths + import torch + import cv2 + from cv2 import setNumThreads + import numpy as np + import comfy.samplers + import comfy.sd + import warnings + from PIL import Image, ImageFilter + from skimage.measure import label, regionprops + from collections import namedtuple + import piexif + + if not impact.config.get_config()['mmdet_skip']: + import mmcv + from mmdet.apis import (inference_detector, init_detector) + from mmdet.evaluation import get_classes +except Exception as e: + import logging + logging.error("[Impact Pack] Failed to import due to several dependencies are missing!!!!") + raise e + + +import impact.impact_server # to load server api + +from .modules.impact.impact_pack import * +from .modules.impact.detectors import * +from .modules.impact.pipe import * +from .modules.impact.logics import * +from .modules.impact.util_nodes import * +from .modules.impact.segs_nodes import * +from .modules.impact.special_samplers import * +from .modules.impact.hf_nodes import * +from .modules.impact.bridge_nodes import * +from .modules.impact.hook_nodes import * +from .modules.impact.animatediff_nodes import * +from .modules.impact.segs_upscaler import * + +import threading + + +threading.Thread(target=impact.wildcards.wildcard_load).start() + + +NODE_CLASS_MAPPINGS = { + "SAMLoader": SAMLoader, + "CLIPSegDetectorProvider": CLIPSegDetectorProvider, + "ONNXDetectorProvider": ONNXDetectorProvider, + + "BitwiseAndMaskForEach": BitwiseAndMaskForEach, + "SubtractMaskForEach": SubtractMaskForEach, + + "DetailerForEach": DetailerForEach, + "DetailerForEachDebug": DetailerForEachTest, + "DetailerForEachPipe": DetailerForEachPipe, + "DetailerForEachDebugPipe": DetailerForEachTestPipe, + "DetailerForEachPipeForAnimateDiff": DetailerForEachPipeForAnimateDiff, + + "SAMDetectorCombined": SAMDetectorCombined, + "SAMDetectorSegmented": SAMDetectorSegmented, + + "FaceDetailer": FaceDetailer, + "FaceDetailerPipe": FaceDetailerPipe, + "MaskDetailerPipe": MaskDetailerPipe, + + "ToDetailerPipe": ToDetailerPipe, + "ToDetailerPipeSDXL": ToDetailerPipeSDXL, + "FromDetailerPipe": FromDetailerPipe, + "FromDetailerPipe_v2": FromDetailerPipe_v2, + "FromDetailerPipeSDXL": FromDetailerPipe_SDXL, + "AnyPipeToBasic": AnyPipeToBasic, + "ToBasicPipe": ToBasicPipe, + "FromBasicPipe": FromBasicPipe, + "FromBasicPipe_v2": FromBasicPipe_v2, + "BasicPipeToDetailerPipe": BasicPipeToDetailerPipe, + "BasicPipeToDetailerPipeSDXL": BasicPipeToDetailerPipeSDXL, + "DetailerPipeToBasicPipe": DetailerPipeToBasicPipe, + "EditBasicPipe": EditBasicPipe, + "EditDetailerPipe": EditDetailerPipe, + "EditDetailerPipeSDXL": EditDetailerPipeSDXL, + + "LatentPixelScale": LatentPixelScale, + "PixelKSampleUpscalerProvider": PixelKSampleUpscalerProvider, + "PixelKSampleUpscalerProviderPipe": PixelKSampleUpscalerProviderPipe, + "IterativeLatentUpscale": IterativeLatentUpscale, + "IterativeImageUpscale": IterativeImageUpscale, + "PixelTiledKSampleUpscalerProvider": PixelTiledKSampleUpscalerProvider, + "PixelTiledKSampleUpscalerProviderPipe": PixelTiledKSampleUpscalerProviderPipe, + "TwoSamplersForMaskUpscalerProvider": TwoSamplersForMaskUpscalerProvider, + "TwoSamplersForMaskUpscalerProviderPipe": TwoSamplersForMaskUpscalerProviderPipe, + + "PixelKSampleHookCombine": PixelKSampleHookCombine, + "DenoiseScheduleHookProvider": DenoiseScheduleHookProvider, + "StepsScheduleHookProvider": StepsScheduleHookProvider, + "CfgScheduleHookProvider": CfgScheduleHookProvider, + "NoiseInjectionHookProvider": NoiseInjectionHookProvider, + "UnsamplerHookProvider": UnsamplerHookProvider, + "CoreMLDetailerHookProvider": CoreMLDetailerHookProvider, + "PreviewDetailerHookProvider": PreviewDetailerHookProvider, + + "DetailerHookCombine": DetailerHookCombine, + "NoiseInjectionDetailerHookProvider": NoiseInjectionDetailerHookProvider, + "UnsamplerDetailerHookProvider": UnsamplerDetailerHookProvider, + "DenoiseSchedulerDetailerHookProvider": DenoiseSchedulerDetailerHookProvider, + "SEGSOrderedFilterDetailerHookProvider": SEGSOrderedFilterDetailerHookProvider, + "SEGSRangeFilterDetailerHookProvider": SEGSRangeFilterDetailerHookProvider, + "SEGSLabelFilterDetailerHookProvider": SEGSLabelFilterDetailerHookProvider, + "VariationNoiseDetailerHookProvider": VariationNoiseDetailerHookProvider, + # "CustomNoiseDetailerHookProvider": CustomNoiseDetailerHookProvider, + + "BitwiseAndMask": BitwiseAndMask, + "SubtractMask": SubtractMask, + "AddMask": AddMask, + "ImpactSegsAndMask": SegsBitwiseAndMask, + "ImpactSegsAndMaskForEach": SegsBitwiseAndMaskForEach, + "EmptySegs": EmptySEGS, + "ImpactFlattenMask": FlattenMask, + + "MediaPipeFaceMeshToSEGS": MediaPipeFaceMeshToSEGS, + "MaskToSEGS": MaskToSEGS, + "MaskToSEGS_for_AnimateDiff": MaskToSEGS_for_AnimateDiff, + "ToBinaryMask": ToBinaryMask, + "MasksToMaskList": MasksToMaskList, + "MaskListToMaskBatch": MaskListToMaskBatch, + "ImageListToImageBatch": ImageListToImageBatch, + "SetDefaultImageForSEGS": DefaultImageForSEGS, + "RemoveImageFromSEGS": RemoveImageFromSEGS, + + "BboxDetectorSEGS": BboxDetectorForEach, + "SegmDetectorSEGS": SegmDetectorForEach, + "ONNXDetectorSEGS": BboxDetectorForEach, + "ImpactSimpleDetectorSEGS_for_AD": SimpleDetectorForAnimateDiff, + "ImpactSimpleDetectorSEGS": SimpleDetectorForEach, + "ImpactSimpleDetectorSEGSPipe": SimpleDetectorForEachPipe, + "ImpactControlNetApplySEGS": ControlNetApplySEGS, + "ImpactControlNetApplyAdvancedSEGS": ControlNetApplyAdvancedSEGS, + "ImpactControlNetClearSEGS": ControlNetClearSEGS, + "ImpactIPAdapterApplySEGS": IPAdapterApplySEGS, + + "ImpactDecomposeSEGS": DecomposeSEGS, + "ImpactAssembleSEGS": AssembleSEGS, + "ImpactFrom_SEG_ELT": From_SEG_ELT, + "ImpactEdit_SEG_ELT": Edit_SEG_ELT, + "ImpactDilate_Mask_SEG_ELT": Dilate_SEG_ELT, + "ImpactDilateMask": DilateMask, + "ImpactGaussianBlurMask": GaussianBlurMask, + "ImpactDilateMaskInSEGS": DilateMaskInSEGS, + "ImpactGaussianBlurMaskInSEGS": GaussianBlurMaskInSEGS, + "ImpactScaleBy_BBOX_SEG_ELT": SEG_ELT_BBOX_ScaleBy, + "ImpactFrom_SEG_ELT_bbox": From_SEG_ELT_bbox, + "ImpactFrom_SEG_ELT_crop_region": From_SEG_ELT_crop_region, + "ImpactCount_Elts_in_SEGS": Count_Elts_in_SEGS, + + "BboxDetectorCombined_v2": BboxDetectorCombined, + "SegmDetectorCombined_v2": SegmDetectorCombined, + "SegsToCombinedMask": SegsToCombinedMask, + + "KSamplerProvider": KSamplerProvider, + "TwoSamplersForMask": TwoSamplersForMask, + "TiledKSamplerProvider": TiledKSamplerProvider, + + "KSamplerAdvancedProvider": KSamplerAdvancedProvider, + "TwoAdvancedSamplersForMask": TwoAdvancedSamplersForMask, + + "ImpactNegativeConditioningPlaceholder": NegativeConditioningPlaceholder, + + "PreviewBridge": PreviewBridge, + "PreviewBridgeLatent": PreviewBridgeLatent, + "ImageSender": ImageSender, + "ImageReceiver": ImageReceiver, + "LatentSender": LatentSender, + "LatentReceiver": LatentReceiver, + "ImageMaskSwitch": ImageMaskSwitch, + "LatentSwitch": GeneralSwitch, + "SEGSSwitch": GeneralSwitch, + "ImpactSwitch": GeneralSwitch, + "ImpactInversedSwitch": GeneralInversedSwitch, + + "ImpactWildcardProcessor": ImpactWildcardProcessor, + "ImpactWildcardEncode": ImpactWildcardEncode, + + "SEGSUpscaler": SEGSUpscaler, + "SEGSUpscalerPipe": SEGSUpscalerPipe, + "SEGSDetailer": SEGSDetailer, + "SEGSPaste": SEGSPaste, + "SEGSPreview": SEGSPreview, + "SEGSPreviewCNet": SEGSPreviewCNet, + "SEGSToImageList": SEGSToImageList, + "ImpactSEGSToMaskList": SEGSToMaskList, + "ImpactSEGSToMaskBatch": SEGSToMaskBatch, + "ImpactSEGSConcat": SEGSConcat, + "ImpactSEGSPicker": SEGSPicker, + "ImpactMakeTileSEGS": MakeTileSEGS, + "ImpactSEGSMerge": SEGSMerge, + + "SEGSDetailerForAnimateDiff": SEGSDetailerForAnimateDiff, + + "ImpactKSamplerBasicPipe": KSamplerBasicPipe, + "ImpactKSamplerAdvancedBasicPipe": KSamplerAdvancedBasicPipe, + + "ReencodeLatent": ReencodeLatent, + "ReencodeLatentPipe": ReencodeLatentPipe, + + "ImpactImageBatchToImageList": ImageBatchToImageList, + "ImpactMakeImageList": MakeImageList, + "ImpactMakeImageBatch": MakeImageBatch, + "ImpactMakeAnyList": MakeAnyList, + "ImpactMakeMaskList": MakeMaskList, + "ImpactMakeMaskBatch": MakeMaskBatch, + + "RegionalSampler": RegionalSampler, + "RegionalSamplerAdvanced": RegionalSamplerAdvanced, + "CombineRegionalPrompts": CombineRegionalPrompts, + "RegionalPrompt": RegionalPrompt, + + "ImpactCombineConditionings": CombineConditionings, + "ImpactConcatConditionings": ConcatConditionings, + + "ImpactSEGSLabelAssign": SEGSLabelAssign, + "ImpactSEGSLabelFilter": SEGSLabelFilter, + "ImpactSEGSRangeFilter": SEGSRangeFilter, + "ImpactSEGSOrderedFilter": SEGSOrderedFilter, + + "ImpactCompare": ImpactCompare, + "ImpactConditionalBranch": ImpactConditionalBranch, + "ImpactConditionalBranchSelMode": ImpactConditionalBranchSelMode, + "ImpactIfNone": ImpactIfNone, + "ImpactConvertDataType": ImpactConvertDataType, + "ImpactLogicalOperators": ImpactLogicalOperators, + "ImpactInt": ImpactInt, + "ImpactFloat": ImpactFloat, + "ImpactBoolean": ImpactBoolean, + "ImpactValueSender": ImpactValueSender, + "ImpactValueReceiver": ImpactValueReceiver, + "ImpactImageInfo": ImpactImageInfo, + "ImpactLatentInfo": ImpactLatentInfo, + "ImpactMinMax": ImpactMinMax, + "ImpactNeg": ImpactNeg, + "ImpactConditionalStopIteration": ImpactConditionalStopIteration, + "ImpactStringSelector": StringSelector, + "StringListToString": StringListToString, + "WildcardPromptFromString": WildcardPromptFromString, + "ImpactExecutionOrderController": ImpactExecutionOrderController, + + "RemoveNoiseMask": RemoveNoiseMask, + + "ImpactLogger": ImpactLogger, + "ImpactDummyInput": ImpactDummyInput, + + "ImpactQueueTrigger": ImpactQueueTrigger, + "ImpactQueueTriggerCountdown": ImpactQueueTriggerCountdown, + "ImpactSetWidgetValue": ImpactSetWidgetValue, + "ImpactNodeSetMuteState": ImpactNodeSetMuteState, + "ImpactControlBridge": ImpactControlBridge, + "ImpactIsNotEmptySEGS": ImpactNotEmptySEGS, + "ImpactSleep": ImpactSleep, + "ImpactRemoteBoolean": ImpactRemoteBoolean, + "ImpactRemoteInt": ImpactRemoteInt, + + "ImpactHFTransformersClassifierProvider": HF_TransformersClassifierProvider, + "ImpactSEGSClassify": SEGS_Classify, + + "ImpactSchedulerAdapter": ImpactSchedulerAdapter, + "GITSSchedulerFuncProvider": GITSSchedulerFuncProvider +} + + +NODE_DISPLAY_NAME_MAPPINGS = { + "SAMLoader": "SAMLoader (Impact)", + + "BboxDetectorSEGS": "BBOX Detector (SEGS)", + "SegmDetectorSEGS": "SEGM Detector (SEGS)", + "ONNXDetectorSEGS": "ONNX Detector (SEGS/legacy) - use BBOXDetector", + "ImpactSimpleDetectorSEGS_for_AD": "Simple Detector for AnimateDiff (SEGS)", + "ImpactSimpleDetectorSEGS": "Simple Detector (SEGS)", + "ImpactSimpleDetectorSEGSPipe": "Simple Detector (SEGS/pipe)", + "ImpactControlNetApplySEGS": "ControlNetApply (SEGS) - DEPRECATED", + "ImpactControlNetApplyAdvancedSEGS": "ControlNetApply (SEGS)", + "ImpactIPAdapterApplySEGS": "IPAdapterApply (SEGS)", + + "BboxDetectorCombined_v2": "BBOX Detector (combined)", + "SegmDetectorCombined_v2": "SEGM Detector (combined)", + "SegsToCombinedMask": "SEGS to MASK (combined)", + "MediaPipeFaceMeshToSEGS": "MediaPipe FaceMesh to SEGS", + "MaskToSEGS": "MASK to SEGS", + "MaskToSEGS_for_AnimateDiff": "MASK to SEGS for AnimateDiff", + "BitwiseAndMaskForEach": "Pixelwise(SEGS & SEGS)", + "SubtractMaskForEach": "Pixelwise(SEGS - SEGS)", + "ImpactSegsAndMask": "Pixelwise(SEGS & MASK)", + "ImpactSegsAndMaskForEach": "Pixelwise(SEGS & MASKS ForEach)", + "BitwiseAndMask": "Pixelwise(MASK & MASK)", + "SubtractMask": "Pixelwise(MASK - MASK)", + "AddMask": "Pixelwise(MASK + MASK)", + "ImpactFlattenMask": "Flatten Mask Batch", + "DetailerForEach": "Detailer (SEGS)", + "DetailerForEachPipe": "Detailer (SEGS/pipe)", + "DetailerForEachDebug": "DetailerDebug (SEGS)", + "DetailerForEachDebugPipe": "DetailerDebug (SEGS/pipe)", + "SEGSDetailerForAnimateDiff": "SEGSDetailer For AnimateDiff (SEGS/pipe)", + "DetailerForEachPipeForAnimateDiff": "Detailer For AnimateDiff (SEGS/pipe)", + "SEGSUpscaler": "Upscaler (SEGS)", + "SEGSUpscalerPipe": "Upscaler (SEGS/pipe)", + + "SAMDetectorCombined": "SAMDetector (combined)", + "SAMDetectorSegmented": "SAMDetector (segmented)", + "FaceDetailerPipe": "FaceDetailer (pipe)", + "MaskDetailerPipe": "MaskDetailer (pipe)", + + "FromDetailerPipeSDXL": "FromDetailer (SDXL/pipe)", + "BasicPipeToDetailerPipeSDXL": "BasicPipe -> DetailerPipe (SDXL)", + "EditDetailerPipeSDXL": "Edit DetailerPipe (SDXL)", + + "BasicPipeToDetailerPipe": "BasicPipe -> DetailerPipe", + "DetailerPipeToBasicPipe": "DetailerPipe -> BasicPipe", + "EditBasicPipe": "Edit BasicPipe", + "EditDetailerPipe": "Edit DetailerPipe", + "AnyPipeToBasic": "Any PIPE -> BasicPipe", + + "LatentPixelScale": "Latent Scale (on Pixel Space)", + "IterativeLatentUpscale": "Iterative Upscale (Latent/on Pixel Space)", + "IterativeImageUpscale": "Iterative Upscale (Image)", + + "TwoSamplersForMaskUpscalerProvider": "TwoSamplersForMask Upscaler Provider", + "TwoSamplersForMaskUpscalerProviderPipe": "TwoSamplersForMask Upscaler Provider (pipe)", + + "ReencodeLatent": "Reencode Latent", + "ReencodeLatentPipe": "Reencode Latent (pipe)", + + "ImpactKSamplerBasicPipe": "KSampler (pipe)", + "ImpactKSamplerAdvancedBasicPipe": "KSampler (Advanced/pipe)", + "ImpactSEGSLabelAssign": "SEGS Assign (label)", + "ImpactSEGSLabelFilter": "SEGS Filter (label)", + "ImpactSEGSRangeFilter": "SEGS Filter (range)", + "ImpactSEGSOrderedFilter": "SEGS Filter (ordered)", + "ImpactSEGSConcat": "SEGS Concat", + "ImpactSEGSToMaskList": "SEGS to Mask List", + "ImpactSEGSToMaskBatch": "SEGS to Mask Batch", + "ImpactSEGSPicker": "Picker (SEGS)", + "ImpactMakeTileSEGS": "Make Tile SEGS", + "ImpactSEGSMerge": "SEGS Merge", + + "ImpactDecomposeSEGS": "Decompose (SEGS)", + "ImpactAssembleSEGS": "Assemble (SEGS)", + "ImpactFrom_SEG_ELT": "From SEG_ELT", + "ImpactEdit_SEG_ELT": "Edit SEG_ELT", + "ImpactFrom_SEG_ELT_bbox": "From SEG_ELT bbox", + "ImpactFrom_SEG_ELT_crop_region": "From SEG_ELT crop_region", + "ImpactDilate_Mask_SEG_ELT": "Dilate Mask (SEG_ELT)", + "ImpactScaleBy_BBOX_SEG_ELT": "ScaleBy BBOX (SEG_ELT)", + "ImpactCount_Elts_in_SEGS": "Count Elts in SEGS", + "ImpactDilateMask": "Dilate Mask", + "ImpactGaussianBlurMask": "Gaussian Blur Mask", + "ImpactDilateMaskInSEGS": "Dilate Mask (SEGS)", + "ImpactGaussianBlurMaskInSEGS": "Gaussian Blur Mask (SEGS)", + + "PreviewBridge": "Preview Bridge (Image)", + "PreviewBridgeLatent": "Preview Bridge (Latent)", + "ImageSender": "Image Sender", + "ImageReceiver": "Image Receiver", + "ImageMaskSwitch": "Switch (images, mask)", + "ImpactSwitch": "Switch (Any)", + "ImpactInversedSwitch": "Inversed Switch (Any)", + "ImpactExecutionOrderController": "Execution Order Controller", + + "MasksToMaskList": "Mask Batch to Mask List", + "MaskListToMaskBatch": "Mask List to Mask Batch", + "ImpactImageBatchToImageList": "Image Batch to Image List", + "ImageListToImageBatch": "Image List to Image Batch", + + "ImpactMakeImageList": "Make Image List", + "ImpactMakeImageBatch": "Make Image Batch", + "ImpactMakeMaskList": "Make Mask List", + "ImpactMakeMaskBatch": "Make Mask Batch", + "ImpactMakeAnyList": "Make List (Any)", + + "ImpactStringSelector": "String Selector", + "StringListToString": "String List to String", + "WildcardPromptFromString": "Wildcard Prompt from String", + "ImpactIsNotEmptySEGS": "SEGS isn't Empty", + "SetDefaultImageForSEGS": "Set Default Image for SEGS", + "RemoveImageFromSEGS": "Remove Image from SEGS", + + "RemoveNoiseMask": "Remove Noise Mask", + + "ImpactCombineConditionings": "Combine Conditionings", + "ImpactConcatConditionings": "Concat Conditionings", + + "ImpactQueueTrigger": "Queue Trigger", + "ImpactQueueTriggerCountdown": "Queue Trigger (Countdown)", + "ImpactSetWidgetValue": "Set Widget Value", + "ImpactNodeSetMuteState": "Set Mute State", + "ImpactControlBridge": "Control Bridge", + "ImpactSleep": "Sleep", + "ImpactRemoteBoolean": "Remote Boolean (on prompt)", + "ImpactRemoteInt": "Remote Int (on prompt)", + + "ImpactHFTransformersClassifierProvider": "HF Transformers Classifier Provider", + "ImpactSEGSClassify": "SEGS Classify", + + "LatentSwitch": "Switch (latent/legacy)", + "SEGSSwitch": "Switch (SEGS/legacy)", + + "SEGSPreviewCNet": "SEGSPreview (CNET Image)", + + "ImpactSchedulerAdapter": "Impact Scheduler Adapter", + "GITSSchedulerFuncProvider": "GITSScheduler Func Provider", + "ImpactNegativeConditioningPlaceholder": "Negative Cond Placeholder" +} + +if not impact.config.get_config()['mmdet_skip']: + from impact.mmdet_nodes import * + import impact.legacy_nodes + NODE_CLASS_MAPPINGS.update({ + "MMDetDetectorProvider": MMDetDetectorProvider, + "MMDetLoader": impact.legacy_nodes.MMDetLoader, + "MaskPainter": impact.legacy_nodes.MaskPainter, + "SegsMaskCombine": impact.legacy_nodes.SegsMaskCombine, + "BboxDetectorForEach": impact.legacy_nodes.BboxDetectorForEach, + "SegmDetectorForEach": impact.legacy_nodes.SegmDetectorForEach, + "BboxDetectorCombined": impact.legacy_nodes.BboxDetectorCombined, + "SegmDetectorCombined": impact.legacy_nodes.SegmDetectorCombined, + }) + + NODE_DISPLAY_NAME_MAPPINGS.update({ + "MaskPainter": "MaskPainter (Deprecated)", + "MMDetLoader": "MMDetLoader (Legacy)", + "SegsMaskCombine": "SegsMaskCombine (Legacy)", + "BboxDetectorForEach": "BboxDetectorForEach (Legacy)", + "SegmDetectorForEach": "SegmDetectorForEach (Legacy)", + "BboxDetectorCombined": "BboxDetectorCombined (Legacy)", + "SegmDetectorCombined": "SegmDetectorCombined (Legacy)", + }) + +try: + import impact.subpack_nodes + + NODE_CLASS_MAPPINGS.update(impact.subpack_nodes.NODE_CLASS_MAPPINGS) + NODE_DISPLAY_NAME_MAPPINGS.update(impact.subpack_nodes.NODE_DISPLAY_NAME_MAPPINGS) +except Exception as e: + print("### ComfyUI-Impact-Pack: (IMPORT FAILED) Subpack\n") + print(" The module at the `custom_nodes/ComfyUI-Impact-Pack/impact_subpack` path appears to be incomplete.") + print(" Recommended to delete the path and restart ComfyUI.") + print(" If the issue persists, please report it to https://github.com/ltdrdata/ComfyUI-Impact-Pack/issues.") + print("\n---------------------------------") + traceback.print_exc() + print("---------------------------------\n") + +# NOTE: Inject directly into EXTENSION_WEB_DIRS instead of WEB_DIRECTORY +# Provide the js path fixed as ComfyUI-Impact-Pack instead of the path name, making it available for external use + +# WEB_DIRECTORY = "js" -- deprecated method +nodes.EXTENSION_WEB_DIRS["ComfyUI-Impact-Pack"] = os.path.join(os.path.dirname(os.path.realpath(__file__)), 'js') + + +__all__ = ['NODE_CLASS_MAPPINGS', 'NODE_DISPLAY_NAME_MAPPINGS'] + + +try: + import cm_global + cm_global.register_extension('ComfyUI-Impact-Pack', + {'version': config.version_code, + 'name': 'Impact Pack', + 'nodes': set(NODE_CLASS_MAPPINGS.keys()), + 'description': 'This extension provides inpainting functionality based on the detector and detailer, along with convenient workflow features like wildcards and logics.', }) +except: + pass diff --git a/ComfyUI-Impact-Pack/__pycache__/__init__.cpython-312.pyc b/ComfyUI-Impact-Pack/__pycache__/__init__.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..b7ea0767f33db2031f9aa85e1576f31a0f7742e3 Binary files /dev/null and b/ComfyUI-Impact-Pack/__pycache__/__init__.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/custom_wildcards/put_wildcards_here b/ComfyUI-Impact-Pack/custom_wildcards/put_wildcards_here new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/ComfyUI-Impact-Pack/disable.py b/ComfyUI-Impact-Pack/disable.py new file mode 100644 index 0000000000000000000000000000000000000000..900d77e81ebe1e6633a113f06d4f2005a9184c1f --- /dev/null +++ b/ComfyUI-Impact-Pack/disable.py @@ -0,0 +1,38 @@ +import os +import sys +import time +import platform +import shutil +import subprocess + +comfy_path = '../..' + +def rmtree(path): + retry_count = 3 + + while True: + try: + retry_count -= 1 + + if platform.system() == "Windows": + subprocess.check_call(['attrib', '-R', path + '\\*', '/S']) + + shutil.rmtree(path) + + return True + + except Exception as ex: + print(f"ex: {ex}") + time.sleep(3) + + if retry_count < 0: + raise ex + + print(f"Uninstall retry({retry_count})") + +js_dest_path = os.path.join(comfy_path, "web", "extensions", "impact-pack") + +if os.path.exists(js_dest_path): + rmtree(js_dest_path) + + diff --git a/ComfyUI-Impact-Pack/impact-pack.ini b/ComfyUI-Impact-Pack/impact-pack.ini new file mode 100644 index 0000000000000000000000000000000000000000..fc018faa534395b1a375b12e2ac21142f09edb97 --- /dev/null +++ b/ComfyUI-Impact-Pack/impact-pack.ini @@ -0,0 +1,8 @@ +[default] +dependency_version = 23 +mmdet_skip = True +sam_editor_cpu = False +sam_editor_model = sam_vit_b_01ec64.pth +custom_wildcards = C:\Users\NATO\Documents\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-Impact-Pack\custom_wildcards +disable_gpu_opencv = True + diff --git a/ComfyUI-Impact-Pack/impact_subpack/.gitignore b/ComfyUI-Impact-Pack/impact_subpack/.gitignore new file mode 100644 index 0000000000000000000000000000000000000000..87bd918e20f98435aaca1749182eda4d6ee94e3a --- /dev/null +++ b/ComfyUI-Impact-Pack/impact_subpack/.gitignore @@ -0,0 +1,5 @@ +__pycache__ +*.ini +wildcards/** +.vscode/ +.idea/ \ No newline at end of file diff --git a/ComfyUI-Impact-Pack/impact_subpack/LICENSE b/ComfyUI-Impact-Pack/impact_subpack/LICENSE new file mode 100644 index 0000000000000000000000000000000000000000..ada1a8176b23b97e70d89300c3f06c3c471bec2c --- /dev/null +++ b/ComfyUI-Impact-Pack/impact_subpack/LICENSE @@ -0,0 +1,661 @@ + GNU AFFERO GENERAL PUBLIC LICENSE + Version 3, 19 November 2007 + + Copyright (C) 2007 Free Software Foundation, Inc. + Everyone is permitted to copy and distribute verbatim copies + of this license document, but changing it is not allowed. + + Preamble + + The GNU Affero General Public License is a free, copyleft license for +software and other kinds of works, specifically designed to ensure +cooperation with the community in the case of network server software. + + The licenses for most software and other practical works are designed +to take away your freedom to share and change the works. By contrast, +our General Public Licenses are intended to guarantee your freedom to +share and change all versions of a program--to make sure it remains free +software for all its users. + + When we speak of free software, we are referring to freedom, not +price. Our General Public Licenses are designed to make sure that you +have the freedom to distribute copies of free software (and charge for +them if you wish), that you receive source code or can get it if you +want it, that you can change the software or use pieces of it in new +free programs, and that you know you can do these things. + + Developers that use our General Public Licenses protect your rights +with two steps: (1) assert copyright on the software, and (2) offer +you this License which gives you legal permission to copy, distribute +and/or modify the software. + + A secondary benefit of defending all users' freedom is that +improvements made in alternate versions of the program, if they +receive widespread use, become available for other developers to +incorporate. Many developers of free software are heartened and +encouraged by the resulting cooperation. However, in the case of +software used on network servers, this result may fail to come about. +The GNU General Public License permits making a modified version and +letting the public access it on a server without ever releasing its +source code to the public. + + The GNU Affero General Public License is designed specifically to +ensure that, in such cases, the modified source code becomes available +to the community. It requires the operator of a network server to +provide the source code of the modified version running there to the +users of that server. Therefore, public use of a modified version, on +a publicly accessible server, gives the public access to the source +code of the modified version. + + An older license, called the Affero General Public License and +published by Affero, was designed to accomplish similar goals. This is +a different license, not a version of the Affero GPL, but Affero has +released a new version of the Affero GPL which permits relicensing under +this license. + + The precise terms and conditions for copying, distribution and +modification follow. + + TERMS AND CONDITIONS + + 0. Definitions. + + "This License" refers to version 3 of the GNU Affero General Public License. + + "Copyright" also means copyright-like laws that apply to other kinds of +works, such as semiconductor masks. + + "The Program" refers to any copyrightable work licensed under this +License. Each licensee is addressed as "you". "Licensees" and +"recipients" may be individuals or organizations. + + To "modify" a work means to copy from or adapt all or part of the work +in a fashion requiring copyright permission, other than the making of an +exact copy. The resulting work is called a "modified version" of the +earlier work or a work "based on" the earlier work. + + A "covered work" means either the unmodified Program or a work based +on the Program. + + To "propagate" a work means to do anything with it that, without +permission, would make you directly or secondarily liable for +infringement under applicable copyright law, except executing it on a +computer or modifying a private copy. Propagation includes copying, +distribution (with or without modification), making available to the +public, and in some countries other activities as well. + + To "convey" a work means any kind of propagation that enables other +parties to make or receive copies. Mere interaction with a user through +a computer network, with no transfer of a copy, is not conveying. + + An interactive user interface displays "Appropriate Legal Notices" +to the extent that it includes a convenient and prominently visible +feature that (1) displays an appropriate copyright notice, and (2) +tells the user that there is no warranty for the work (except to the +extent that warranties are provided), that licensees may convey the +work under this License, and how to view a copy of this License. If +the interface presents a list of user commands or options, such as a +menu, a prominent item in the list meets this criterion. + + 1. Source Code. + + The "source code" for a work means the preferred form of the work +for making modifications to it. "Object code" means any non-source +form of a work. + + A "Standard Interface" means an interface that either is an official +standard defined by a recognized standards body, or, in the case of +interfaces specified for a particular programming language, one that +is widely used among developers working in that language. + + The "System Libraries" of an executable work include anything, other +than the work as a whole, that (a) is included in the normal form of +packaging a Major Component, but which is not part of that Major +Component, and (b) serves only to enable use of the work with that +Major Component, or to implement a Standard Interface for which an +implementation is available to the public in source code form. A +"Major Component", in this context, means a major essential component +(kernel, window system, and so on) of the specific operating system +(if any) on which the executable work runs, or a compiler used to +produce the work, or an object code interpreter used to run it. + + The "Corresponding Source" for a work in object code form means all +the source code needed to generate, install, and (for an executable +work) run the object code and to modify the work, including scripts to +control those activities. However, it does not include the work's +System Libraries, or general-purpose tools or generally available free +programs which are used unmodified in performing those activities but +which are not part of the work. For example, Corresponding Source +includes interface definition files associated with source files for +the work, and the source code for shared libraries and dynamically +linked subprograms that the work is specifically designed to require, +such as by intimate data communication or control flow between those +subprograms and other parts of the work. + + The Corresponding Source need not include anything that users +can regenerate automatically from other parts of the Corresponding +Source. + + The Corresponding Source for a work in source code form is that +same work. + + 2. Basic Permissions. + + All rights granted under this License are granted for the term of +copyright on the Program, and are irrevocable provided the stated +conditions are met. This License explicitly affirms your unlimited +permission to run the unmodified Program. The output from running a +covered work is covered by this License only if the output, given its +content, constitutes a covered work. This License acknowledges your +rights of fair use or other equivalent, as provided by copyright law. + + You may make, run and propagate covered works that you do not +convey, without conditions so long as your license otherwise remains +in force. You may convey covered works to others for the sole purpose +of having them make modifications exclusively for you, or provide you +with facilities for running those works, provided that you comply with +the terms of this License in conveying all material for which you do +not control copyright. Those thus making or running the covered works +for you must do so exclusively on your behalf, under your direction +and control, on terms that prohibit them from making any copies of +your copyrighted material outside their relationship with you. + + Conveying under any other circumstances is permitted solely under +the conditions stated below. Sublicensing is not allowed; section 10 +makes it unnecessary. + + 3. Protecting Users' Legal Rights From Anti-Circumvention Law. + + No covered work shall be deemed part of an effective technological +measure under any applicable law fulfilling obligations under article +11 of the WIPO copyright treaty adopted on 20 December 1996, or +similar laws prohibiting or restricting circumvention of such +measures. + + When you convey a covered work, you waive any legal power to forbid +circumvention of technological measures to the extent such circumvention +is effected by exercising rights under this License with respect to +the covered work, and you disclaim any intention to limit operation or +modification of the work as a means of enforcing, against the work's +users, your or third parties' legal rights to forbid circumvention of +technological measures. + + 4. Conveying Verbatim Copies. + + You may convey verbatim copies of the Program's source code as you +receive it, in any medium, provided that you conspicuously and +appropriately publish on each copy an appropriate copyright notice; +keep intact all notices stating that this License and any +non-permissive terms added in accord with section 7 apply to the code; +keep intact all notices of the absence of any warranty; and give all +recipients a copy of this License along with the Program. + + You may charge any price or no price for each copy that you convey, +and you may offer support or warranty protection for a fee. + + 5. Conveying Modified Source Versions. + + You may convey a work based on the Program, or the modifications to +produce it from the Program, in the form of source code under the +terms of section 4, provided that you also meet all of these conditions: + + a) The work must carry prominent notices stating that you modified + it, and giving a relevant date. + + b) The work must carry prominent notices stating that it is + released under this License and any conditions added under section + 7. This requirement modifies the requirement in section 4 to + "keep intact all notices". + + c) You must license the entire work, as a whole, under this + License to anyone who comes into possession of a copy. This + License will therefore apply, along with any applicable section 7 + additional terms, to the whole of the work, and all its parts, + regardless of how they are packaged. This License gives no + permission to license the work in any other way, but it does not + invalidate such permission if you have separately received it. + + d) If the work has interactive user interfaces, each must display + Appropriate Legal Notices; however, if the Program has interactive + interfaces that do not display Appropriate Legal Notices, your + work need not make them do so. + + A compilation of a covered work with other separate and independent +works, which are not by their nature extensions of the covered work, +and which are not combined with it such as to form a larger program, +in or on a volume of a storage or distribution medium, is called an +"aggregate" if the compilation and its resulting copyright are not +used to limit the access or legal rights of the compilation's users +beyond what the individual works permit. Inclusion of a covered work +in an aggregate does not cause this License to apply to the other +parts of the aggregate. + + 6. Conveying Non-Source Forms. + + You may convey a covered work in object code form under the terms +of sections 4 and 5, provided that you also convey the +machine-readable Corresponding Source under the terms of this License, +in one of these ways: + + a) Convey the object code in, or embodied in, a physical product + (including a physical distribution medium), accompanied by the + Corresponding Source fixed on a durable physical medium + customarily used for software interchange. + + b) Convey the object code in, or embodied in, a physical product + (including a physical distribution medium), accompanied by a + written offer, valid for at least three years and valid for as + long as you offer spare parts or customer support for that product + model, to give anyone who possesses the object code either (1) a + copy of the Corresponding Source for all the software in the + product that is covered by this License, on a durable physical + medium customarily used for software interchange, for a price no + more than your reasonable cost of physically performing this + conveying of source, or (2) access to copy the + Corresponding Source from a network server at no charge. + + c) Convey individual copies of the object code with a copy of the + written offer to provide the Corresponding Source. This + alternative is allowed only occasionally and noncommercially, and + only if you received the object code with such an offer, in accord + with subsection 6b. + + d) Convey the object code by offering access from a designated + place (gratis or for a charge), and offer equivalent access to the + Corresponding Source in the same way through the same place at no + further charge. You need not require recipients to copy the + Corresponding Source along with the object code. If the place to + copy the object code is a network server, the Corresponding Source + may be on a different server (operated by you or a third party) + that supports equivalent copying facilities, provided you maintain + clear directions next to the object code saying where to find the + Corresponding Source. Regardless of what server hosts the + Corresponding Source, you remain obligated to ensure that it is + available for as long as needed to satisfy these requirements. + + e) Convey the object code using peer-to-peer transmission, provided + you inform other peers where the object code and Corresponding + Source of the work are being offered to the general public at no + charge under subsection 6d. + + A separable portion of the object code, whose source code is excluded +from the Corresponding Source as a System Library, need not be +included in conveying the object code work. + + A "User Product" is either (1) a "consumer product", which means any +tangible personal property which is normally used for personal, family, +or household purposes, or (2) anything designed or sold for incorporation +into a dwelling. In determining whether a product is a consumer product, +doubtful cases shall be resolved in favor of coverage. For a particular +product received by a particular user, "normally used" refers to a +typical or common use of that class of product, regardless of the status +of the particular user or of the way in which the particular user +actually uses, or expects or is expected to use, the product. A product +is a consumer product regardless of whether the product has substantial +commercial, industrial or non-consumer uses, unless such uses represent +the only significant mode of use of the product. + + "Installation Information" for a User Product means any methods, +procedures, authorization keys, or other information required to install +and execute modified versions of a covered work in that User Product from +a modified version of its Corresponding Source. The information must +suffice to ensure that the continued functioning of the modified object +code is in no case prevented or interfered with solely because +modification has been made. + + If you convey an object code work under this section in, or with, or +specifically for use in, a User Product, and the conveying occurs as +part of a transaction in which the right of possession and use of the +User Product is transferred to the recipient in perpetuity or for a +fixed term (regardless of how the transaction is characterized), the +Corresponding Source conveyed under this section must be accompanied +by the Installation Information. But this requirement does not apply +if neither you nor any third party retains the ability to install +modified object code on the User Product (for example, the work has +been installed in ROM). + + The requirement to provide Installation Information does not include a +requirement to continue to provide support service, warranty, or updates +for a work that has been modified or installed by the recipient, or for +the User Product in which it has been modified or installed. Access to a +network may be denied when the modification itself materially and +adversely affects the operation of the network or violates the rules and +protocols for communication across the network. + + Corresponding Source conveyed, and Installation Information provided, +in accord with this section must be in a format that is publicly +documented (and with an implementation available to the public in +source code form), and must require no special password or key for +unpacking, reading or copying. + + 7. Additional Terms. + + "Additional permissions" are terms that supplement the terms of this +License by making exceptions from one or more of its conditions. +Additional permissions that are applicable to the entire Program shall +be treated as though they were included in this License, to the extent +that they are valid under applicable law. If additional permissions +apply only to part of the Program, that part may be used separately +under those permissions, but the entire Program remains governed by +this License without regard to the additional permissions. + + When you convey a copy of a covered work, you may at your option +remove any additional permissions from that copy, or from any part of +it. (Additional permissions may be written to require their own +removal in certain cases when you modify the work.) You may place +additional permissions on material, added by you to a covered work, +for which you have or can give appropriate copyright permission. + + Notwithstanding any other provision of this License, for material you +add to a covered work, you may (if authorized by the copyright holders of +that material) supplement the terms of this License with terms: + + a) Disclaiming warranty or limiting liability differently from the + terms of sections 15 and 16 of this License; or + + b) Requiring preservation of specified reasonable legal notices or + author attributions in that material or in the Appropriate Legal + Notices displayed by works containing it; or + + c) Prohibiting misrepresentation of the origin of that material, or + requiring that modified versions of such material be marked in + reasonable ways as different from the original version; or + + d) Limiting the use for publicity purposes of names of licensors or + authors of the material; or + + e) Declining to grant rights under trademark law for use of some + trade names, trademarks, or service marks; or + + f) Requiring indemnification of licensors and authors of that + material by anyone who conveys the material (or modified versions of + it) with contractual assumptions of liability to the recipient, for + any liability that these contractual assumptions directly impose on + those licensors and authors. + + All other non-permissive additional terms are considered "further +restrictions" within the meaning of section 10. If the Program as you +received it, or any part of it, contains a notice stating that it is +governed by this License along with a term that is a further +restriction, you may remove that term. If a license document contains +a further restriction but permits relicensing or conveying under this +License, you may add to a covered work material governed by the terms +of that license document, provided that the further restriction does +not survive such relicensing or conveying. + + If you add terms to a covered work in accord with this section, you +must place, in the relevant source files, a statement of the +additional terms that apply to those files, or a notice indicating +where to find the applicable terms. + + Additional terms, permissive or non-permissive, may be stated in the +form of a separately written license, or stated as exceptions; +the above requirements apply either way. + + 8. Termination. + + You may not propagate or modify a covered work except as expressly +provided under this License. Any attempt otherwise to propagate or +modify it is void, and will automatically terminate your rights under +this License (including any patent licenses granted under the third +paragraph of section 11). + + However, if you cease all violation of this License, then your +license from a particular copyright holder is reinstated (a) +provisionally, unless and until the copyright holder explicitly and +finally terminates your license, and (b) permanently, if the copyright +holder fails to notify you of the violation by some reasonable means +prior to 60 days after the cessation. + + Moreover, your license from a particular copyright holder is +reinstated permanently if the copyright holder notifies you of the +violation by some reasonable means, this is the first time you have +received notice of violation of this License (for any work) from that +copyright holder, and you cure the violation prior to 30 days after +your receipt of the notice. + + Termination of your rights under this section does not terminate the +licenses of parties who have received copies or rights from you under +this License. If your rights have been terminated and not permanently +reinstated, you do not qualify to receive new licenses for the same +material under section 10. + + 9. Acceptance Not Required for Having Copies. + + You are not required to accept this License in order to receive or +run a copy of the Program. Ancillary propagation of a covered work +occurring solely as a consequence of using peer-to-peer transmission +to receive a copy likewise does not require acceptance. However, +nothing other than this License grants you permission to propagate or +modify any covered work. These actions infringe copyright if you do +not accept this License. Therefore, by modifying or propagating a +covered work, you indicate your acceptance of this License to do so. + + 10. Automatic Licensing of Downstream Recipients. + + Each time you convey a covered work, the recipient automatically +receives a license from the original licensors, to run, modify and +propagate that work, subject to this License. You are not responsible +for enforcing compliance by third parties with this License. + + An "entity transaction" is a transaction transferring control of an +organization, or substantially all assets of one, or subdividing an +organization, or merging organizations. If propagation of a covered +work results from an entity transaction, each party to that +transaction who receives a copy of the work also receives whatever +licenses to the work the party's predecessor in interest had or could +give under the previous paragraph, plus a right to possession of the +Corresponding Source of the work from the predecessor in interest, if +the predecessor has it or can get it with reasonable efforts. + + You may not impose any further restrictions on the exercise of the +rights granted or affirmed under this License. For example, you may +not impose a license fee, royalty, or other charge for exercise of +rights granted under this License, and you may not initiate litigation +(including a cross-claim or counterclaim in a lawsuit) alleging that +any patent claim is infringed by making, using, selling, offering for +sale, or importing the Program or any portion of it. + + 11. Patents. + + A "contributor" is a copyright holder who authorizes use under this +License of the Program or a work on which the Program is based. The +work thus licensed is called the contributor's "contributor version". + + A contributor's "essential patent claims" are all patent claims +owned or controlled by the contributor, whether already acquired or +hereafter acquired, that would be infringed by some manner, permitted +by this License, of making, using, or selling its contributor version, +but do not include claims that would be infringed only as a +consequence of further modification of the contributor version. For +purposes of this definition, "control" includes the right to grant +patent sublicenses in a manner consistent with the requirements of +this License. + + Each contributor grants you a non-exclusive, worldwide, royalty-free +patent license under the contributor's essential patent claims, to +make, use, sell, offer for sale, import and otherwise run, modify and +propagate the contents of its contributor version. + + In the following three paragraphs, a "patent license" is any express +agreement or commitment, however denominated, not to enforce a patent +(such as an express permission to practice a patent or covenant not to +sue for patent infringement). To "grant" such a patent license to a +party means to make such an agreement or commitment not to enforce a +patent against the party. + + If you convey a covered work, knowingly relying on a patent license, +and the Corresponding Source of the work is not available for anyone +to copy, free of charge and under the terms of this License, through a +publicly available network server or other readily accessible means, +then you must either (1) cause the Corresponding Source to be so +available, or (2) arrange to deprive yourself of the benefit of the +patent license for this particular work, or (3) arrange, in a manner +consistent with the requirements of this License, to extend the patent +license to downstream recipients. "Knowingly relying" means you have +actual knowledge that, but for the patent license, your conveying the +covered work in a country, or your recipient's use of the covered work +in a country, would infringe one or more identifiable patents in that +country that you have reason to believe are valid. + + If, pursuant to or in connection with a single transaction or +arrangement, you convey, or propagate by procuring conveyance of, a +covered work, and grant a patent license to some of the parties +receiving the covered work authorizing them to use, propagate, modify +or convey a specific copy of the covered work, then the patent license +you grant is automatically extended to all recipients of the covered +work and works based on it. + + A patent license is "discriminatory" if it does not include within +the scope of its coverage, prohibits the exercise of, or is +conditioned on the non-exercise of one or more of the rights that are +specifically granted under this License. You may not convey a covered +work if you are a party to an arrangement with a third party that is +in the business of distributing software, under which you make payment +to the third party based on the extent of your activity of conveying +the work, and under which the third party grants, to any of the +parties who would receive the covered work from you, a discriminatory +patent license (a) in connection with copies of the covered work +conveyed by you (or copies made from those copies), or (b) primarily +for and in connection with specific products or compilations that +contain the covered work, unless you entered into that arrangement, +or that patent license was granted, prior to 28 March 2007. + + Nothing in this License shall be construed as excluding or limiting +any implied license or other defenses to infringement that may +otherwise be available to you under applicable patent law. + + 12. No Surrender of Others' Freedom. + + If conditions are imposed on you (whether by court order, agreement or +otherwise) that contradict the conditions of this License, they do not +excuse you from the conditions of this License. If you cannot convey a +covered work so as to satisfy simultaneously your obligations under this +License and any other pertinent obligations, then as a consequence you may +not convey it at all. For example, if you agree to terms that obligate you +to collect a royalty for further conveying from those to whom you convey +the Program, the only way you could satisfy both those terms and this +License would be to refrain entirely from conveying the Program. + + 13. Remote Network Interaction; Use with the GNU General Public License. + + Notwithstanding any other provision of this License, if you modify the +Program, your modified version must prominently offer all users +interacting with it remotely through a computer network (if your version +supports such interaction) an opportunity to receive the Corresponding +Source of your version by providing access to the Corresponding Source +from a network server at no charge, through some standard or customary +means of facilitating copying of software. This Corresponding Source +shall include the Corresponding Source for any work covered by version 3 +of the GNU General Public License that is incorporated pursuant to the +following paragraph. + + Notwithstanding any other provision of this License, you have +permission to link or combine any covered work with a work licensed +under version 3 of the GNU General Public License into a single +combined work, and to convey the resulting work. The terms of this +License will continue to apply to the part which is the covered work, +but the work with which it is combined will remain governed by version +3 of the GNU General Public License. + + 14. Revised Versions of this License. + + The Free Software Foundation may publish revised and/or new versions of +the GNU Affero General Public License from time to time. Such new versions +will be similar in spirit to the present version, but may differ in detail to +address new problems or concerns. + + Each version is given a distinguishing version number. If the +Program specifies that a certain numbered version of the GNU Affero General +Public License "or any later version" applies to it, you have the +option of following the terms and conditions either of that numbered +version or of any later version published by the Free Software +Foundation. If the Program does not specify a version number of the +GNU Affero General Public License, you may choose any version ever published +by the Free Software Foundation. + + If the Program specifies that a proxy can decide which future +versions of the GNU Affero General Public License can be used, that proxy's +public statement of acceptance of a version permanently authorizes you +to choose that version for the Program. + + Later license versions may give you additional or different +permissions. However, no additional obligations are imposed on any +author or copyright holder as a result of your choosing to follow a +later version. + + 15. Disclaimer of Warranty. + + THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY +APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT +HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY +OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, +THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR +PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM +IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF +ALL NECESSARY SERVICING, REPAIR OR CORRECTION. + + 16. Limitation of Liability. + + IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING +WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS +THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY +GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE +USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF +DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD +PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), +EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF +SUCH DAMAGES. + + 17. Interpretation of Sections 15 and 16. + + If the disclaimer of warranty and limitation of liability provided +above cannot be given local legal effect according to their terms, +reviewing courts shall apply local law that most closely approximates +an absolute waiver of all civil liability in connection with the +Program, unless a warranty or assumption of liability accompanies a +copy of the Program in return for a fee. + + END OF TERMS AND CONDITIONS + + How to Apply These Terms to Your New Programs + + If you develop a new program, and you want it to be of the greatest +possible use to the public, the best way to achieve this is to make it +free software which everyone can redistribute and change under these terms. + + To do so, attach the following notices to the program. It is safest +to attach them to the start of each source file to most effectively +state the exclusion of warranty; and each file should have at least +the "copyright" line and a pointer to where the full notice is found. + + + Copyright (C) + + This program is free software: you can redistribute it and/or modify + it under the terms of the GNU Affero General Public License as published + by the Free Software Foundation, either version 3 of the License, or + (at your option) any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU Affero General Public License for more details. + + You should have received a copy of the GNU Affero General Public License + along with this program. If not, see . + +Also add information on how to contact you by electronic and paper mail. + + If your software can interact with users remotely through a computer +network, you should also make sure that it provides a way for users to +get its source. For example, if your program is a web application, its +interface could display a "Source" link that leads users to an archive +of the code. There are many ways you could offer source, and different +solutions will be better for different programs; see section 13 for the +specific requirements. + + You should also get your employer (if you work as a programmer) or school, +if any, to sign a "copyright disclaimer" for the program, if necessary. +For more information on this, and how to apply and follow the GNU AGPL, see +. diff --git a/ComfyUI-Impact-Pack/impact_subpack/README.md b/ComfyUI-Impact-Pack/impact_subpack/README.md new file mode 100644 index 0000000000000000000000000000000000000000..2f605b1f61b5f53f12f0f02c16297ace09ef38d9 --- /dev/null +++ b/ComfyUI-Impact-Pack/impact_subpack/README.md @@ -0,0 +1,18 @@ +# ComfyUI-Impact-Subpack +This extension serves as a complement to the Impact Pack, offering features that are not deemed suitable for inclusion by default in the ComfyUI Impact Pack. + +The nodes in this repository cannot be used standalone and depend on [ComfyUI-Impact-Pack](https://github.com/ltdrdata/ComfyUI-Impact-Pack). + +## Nodes +* UltralyticsDetectorProvider - This node provides an object detection detector based on Ultralystics. + * By using this Detector Provider, you can replace the existing mmdet-based detector. + + +## Credits + +ComfyUI/[ComfyUI](https://github.com/comfyanonymous/ComfyUI) - A powerful and modular stable diffusion GUI. + +Bing-su/[adetailer](https://github.com/Bing-su/adetailer/) - This repository provides an object detection model and features based on Ultralystics. + +huggingface/Bingsu/[adetailer](https://huggingface.co/Bingsu/adetailer/tree/main) - This repository offers various models based on Ultralystics. +* You can download other models supported by the UltralyticsDetectorProvider from here. diff --git a/ComfyUI-Impact-Pack/impact_subpack/impact/__pycache__/subcore.cpython-312.pyc b/ComfyUI-Impact-Pack/impact_subpack/impact/__pycache__/subcore.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..265a9000dad23cb647fbdcc9f532f950c42c945c Binary files /dev/null and b/ComfyUI-Impact-Pack/impact_subpack/impact/__pycache__/subcore.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/impact_subpack/impact/__pycache__/subpack_nodes.cpython-312.pyc b/ComfyUI-Impact-Pack/impact_subpack/impact/__pycache__/subpack_nodes.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..e0745ab615f34de1c570bcd7905ca2d8e145858e Binary files /dev/null and b/ComfyUI-Impact-Pack/impact_subpack/impact/__pycache__/subpack_nodes.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/impact_subpack/impact/subcore.py b/ComfyUI-Impact-Pack/impact_subpack/impact/subcore.py new file mode 100644 index 0000000000000000000000000000000000000000..d9e0f8d3891dea7e0c49daa91c3cd4c38c0f7e76 --- /dev/null +++ b/ComfyUI-Impact-Pack/impact_subpack/impact/subcore.py @@ -0,0 +1,239 @@ +from pathlib import Path +from PIL import Image + +import impact.core as core +import cv2 +import numpy as np +from torchvision.transforms.functional import to_pil_image +import torch + +orig_torch_load = torch.load + +try: + from ultralytics import YOLO +except Exception as e: + print(e) + print(f"\n!!!!!\n\n[ComfyUI-Impact-Subpack] If this error occurs, please check the following link:\n\thttps://github.com/ltdrdata/ComfyUI-Impact-Pack/blob/Main/troubleshooting/TROUBLESHOOTING.md\n\n!!!!!\n") + raise e + +# HOTFIX: https://github.com/ltdrdata/ComfyUI-Impact-Pack/issues/754 +# importing YOLO breaking original torch.load capabilities +torch.load = orig_torch_load + +def load_yolo(model_path: str): + try: + return YOLO(model_path) + except ModuleNotFoundError: + # https://github.com/ultralytics/ultralytics/issues/3856 + YOLO("yolov8n.pt") + return YOLO(model_path) + + +def inference_bbox( + model, + image: Image.Image, + confidence: float = 0.3, + device: str = "", +): + pred = model(image, conf=confidence, device=device) + + bboxes = pred[0].boxes.xyxy.cpu().numpy() + cv2_image = np.array(image) + if len(cv2_image.shape) == 3: + cv2_image = cv2_image[:, :, ::-1].copy() # Convert RGB to BGR for cv2 processing + else: + # Handle the grayscale image here + # For example, you might want to convert it to a 3-channel grayscale image for consistency: + cv2_image = cv2.cvtColor(cv2_image, cv2.COLOR_GRAY2BGR) + cv2_gray = cv2.cvtColor(cv2_image, cv2.COLOR_BGR2GRAY) + + segms = [] + for x0, y0, x1, y1 in bboxes: + cv2_mask = np.zeros(cv2_gray.shape, np.uint8) + cv2.rectangle(cv2_mask, (int(x0), int(y0)), (int(x1), int(y1)), 255, -1) + cv2_mask_bool = cv2_mask.astype(bool) + segms.append(cv2_mask_bool) + + n, m = bboxes.shape + if n == 0: + return [[], [], [], []] + + results = [[], [], [], []] + for i in range(len(bboxes)): + results[0].append(pred[0].names[int(pred[0].boxes[i].cls.item())]) + results[1].append(bboxes[i]) + results[2].append(segms[i]) + results[3].append(pred[0].boxes[i].conf.cpu().numpy()) + + return results + + +def inference_segm( + model, + image: Image.Image, + confidence: float = 0.3, + device: str = "", +): + pred = model(image, conf=confidence, device=device) + + bboxes = pred[0].boxes.xyxy.cpu().numpy() + n, m = bboxes.shape + if n == 0: + return [[], [], [], []] + + # NOTE: masks.data will be None when n == 0 + segms = pred[0].masks.data.cpu().numpy() + + h_segms = segms.shape[1] + w_segms = segms.shape[2] + h_orig = image.size[1] + w_orig = image.size[0] + ratio_segms = h_segms / w_segms + ratio_orig = h_orig / w_orig + + if ratio_segms == ratio_orig: + h_gap = 0 + w_gap = 0 + elif ratio_segms > ratio_orig: + h_gap = int((ratio_segms - ratio_orig) * h_segms) + w_gap = 0 + else: + h_gap = 0 + ratio_segms = w_segms / h_segms + ratio_orig = w_orig / h_orig + w_gap = int((ratio_segms - ratio_orig) * w_segms) + + results = [[], [], [], []] + for i in range(len(bboxes)): + results[0].append(pred[0].names[int(pred[0].boxes[i].cls.item())]) + results[1].append(bboxes[i]) + + mask = torch.from_numpy(segms[i]) + mask = mask[h_gap:mask.shape[0] - h_gap, w_gap:mask.shape[1] - w_gap] + + scaled_mask = torch.nn.functional.interpolate(mask.unsqueeze(0).unsqueeze(0), size=(image.size[1], image.size[0]), + mode='bilinear', align_corners=False) + scaled_mask = scaled_mask.squeeze().squeeze() + + results[2].append(scaled_mask.numpy()) + results[3].append(pred[0].boxes[i].conf.cpu().numpy()) + + return results + + +class UltraBBoxDetector: + bbox_model = None + + def __init__(self, bbox_model): + self.bbox_model = bbox_model + + def detect(self, image, threshold, dilation, crop_factor, drop_size=1, detailer_hook=None): + drop_size = max(drop_size, 1) + detected_results = inference_bbox(self.bbox_model, core.tensor2pil(image), threshold) + segmasks = core.create_segmasks(detected_results) + + if dilation > 0: + segmasks = core.dilate_masks(segmasks, dilation) + + items = [] + h = image.shape[1] + w = image.shape[2] + + for x, label in zip(segmasks, detected_results[0]): + item_bbox = x[0] + item_mask = x[1] + + y1, x1, y2, x2 = item_bbox + + if x2 - x1 > drop_size and y2 - y1 > drop_size: # minimum dimension must be (2,2) to avoid squeeze issue + crop_region = core.make_crop_region(w, h, item_bbox, crop_factor) + + if detailer_hook is not None: + crop_region = detailer_hook.post_crop_region(w, h, item_bbox, crop_region) + + cropped_image = core.crop_image(image, crop_region) + cropped_mask = core.crop_ndarray2(item_mask, crop_region) + confidence = x[2] + # bbox_size = (item_bbox[2]-item_bbox[0],item_bbox[3]-item_bbox[1]) # (w,h) + + item = core.SEG(cropped_image, cropped_mask, confidence, crop_region, item_bbox, label, None) + + items.append(item) + + shape = image.shape[1], image.shape[2] + segs = shape, items + + if detailer_hook is not None and hasattr(detailer_hook, "post_detection"): + segs = detailer_hook.post_detection(segs) + + return segs + + def detect_combined(self, image, threshold, dilation): + detected_results = inference_bbox(self.bbox_model, core.tensor2pil(image), threshold) + segmasks = core.create_segmasks(detected_results) + if dilation > 0: + segmasks = core.dilate_masks(segmasks, dilation) + + return core.combine_masks(segmasks) + + def setAux(self, x): + pass + + +class UltraSegmDetector: + bbox_model = None + + def __init__(self, bbox_model): + self.bbox_model = bbox_model + + def detect(self, image, threshold, dilation, crop_factor, drop_size=1, detailer_hook=None): + drop_size = max(drop_size, 1) + detected_results = inference_segm(self.bbox_model, core.tensor2pil(image), threshold) + segmasks = core.create_segmasks(detected_results) + + if dilation > 0: + segmasks = core.dilate_masks(segmasks, dilation) + + items = [] + h = image.shape[1] + w = image.shape[2] + + for x, label in zip(segmasks, detected_results[0]): + item_bbox = x[0] + item_mask = x[1] + + y1, x1, y2, x2 = item_bbox + + if x2 - x1 > drop_size and y2 - y1 > drop_size: # minimum dimension must be (2,2) to avoid squeeze issue + crop_region = core.make_crop_region(w, h, item_bbox, crop_factor) + + if detailer_hook is not None: + crop_region = detailer_hook.post_crop_region(w, h, item_bbox, crop_region) + + cropped_image = core.crop_image(image, crop_region) + cropped_mask = core.crop_ndarray2(item_mask, crop_region) + confidence = x[2] + # bbox_size = (item_bbox[2]-item_bbox[0],item_bbox[3]-item_bbox[1]) # (w,h) + + item = core.SEG(cropped_image, cropped_mask, confidence, crop_region, item_bbox, label, None) + + items.append(item) + + shape = image.shape[1], image.shape[2] + segs = shape, items + + if detailer_hook is not None and hasattr(detailer_hook, "post_detection"): + segs = detailer_hook.post_detection(segs) + + return segs + + def detect_combined(self, image, threshold, dilation): + detected_results = inference_segm(self.bbox_model, core.tensor2pil(image), threshold) + segmasks = core.create_segmasks(detected_results) + if dilation > 0: + segmasks = core.dilate_masks(segmasks, dilation) + + return core.combine_masks(segmasks) + + def setAux(self, x): + pass \ No newline at end of file diff --git a/ComfyUI-Impact-Pack/impact_subpack/impact/subpack_nodes.py b/ComfyUI-Impact-Pack/impact_subpack/impact/subpack_nodes.py new file mode 100644 index 0000000000000000000000000000000000000000..e3869b0516aae7bd33cb18c9020aabf4d9fa73d7 --- /dev/null +++ b/ComfyUI-Impact-Pack/impact_subpack/impact/subpack_nodes.py @@ -0,0 +1,45 @@ +import os +import folder_paths +import impact.core as core +import impact.subcore as subcore +from impact.utils import add_folder_path_and_extensions + +version_code = 23 + +print(f"### Loading: ComfyUI-Impact-Pack (Subpack: V0.7)") + +model_path = folder_paths.models_dir +add_folder_path_and_extensions("ultralytics_bbox", [os.path.join(model_path, "ultralytics", "bbox")], folder_paths.supported_pt_extensions) +add_folder_path_and_extensions("ultralytics_segm", [os.path.join(model_path, "ultralytics", "segm")], folder_paths.supported_pt_extensions) +add_folder_path_and_extensions("ultralytics", [os.path.join(model_path, "ultralytics")], folder_paths.supported_pt_extensions) + + +class UltralyticsDetectorProvider: + @classmethod + def INPUT_TYPES(s): + bboxs = ["bbox/"+x for x in folder_paths.get_filename_list("ultralytics_bbox")] + segms = ["segm/"+x for x in folder_paths.get_filename_list("ultralytics_segm")] + return {"required": {"model_name": (bboxs + segms, )}} + RETURN_TYPES = ("BBOX_DETECTOR", "SEGM_DETECTOR") + FUNCTION = "doit" + + CATEGORY = "ImpactPack" + + def doit(self, model_name): + model_path = folder_paths.get_full_path("ultralytics", model_name) + model = subcore.load_yolo(model_path) + + if model_name.startswith("bbox"): + return subcore.UltraBBoxDetector(model), core.NO_SEGM_DETECTOR() + else: + return subcore.UltraBBoxDetector(model), subcore.UltraSegmDetector(model) + + +NODE_CLASS_MAPPINGS = { + "UltralyticsDetectorProvider": UltralyticsDetectorProvider +} + + +NODE_DISPLAY_NAME_MAPPINGS = { + +} diff --git a/ComfyUI-Impact-Pack/impact_subpack/install.py b/ComfyUI-Impact-Pack/impact_subpack/install.py new file mode 100644 index 0000000000000000000000000000000000000000..19756b2018816490d64851b57acff4078a5d0665 --- /dev/null +++ b/ComfyUI-Impact-Pack/impact_subpack/install.py @@ -0,0 +1,39 @@ +import os +import sys +from torchvision.datasets.utils import download_url + +subpack_path = os.path.join(os.path.dirname(__file__)) + +comfy_path = os.environ.get('COMFYUI_PATH') +if comfy_path is None: + print(f"\n[bold yellow]WARN: The `COMFYUI_PATH` environment variable is not set. Assuming `{os.path.dirname(__file__)}/../../../` as the ComfyUI path.[/bold yellow]", file=sys.stderr) + comfy_path = os.path.abspath(os.path.join(os.path.dirname(__file__), '..', '..', '..')) + +sys.path.append(comfy_path) + +model_path = os.environ.get('COMFYUI_MODEL_PATH') +if model_path is None: + print(f"\n[bold yellow]WARN: The `COMFYUI_MODEL_PATH` environment variable is not set. Assuming `{model_path}` as the ComfyUI path.[/bold yellow]", file=sys.stderr) + model_path = os.path.abspath(os.path.join(comfy_path, 'models')) + +ultralytics_bbox_path = os.path.join(model_path, "ultralytics", "bbox") +ultralytics_segm_path = os.path.join(model_path, "ultralytics", "segm") + +if not os.path.exists(os.path.join(subpack_path, '..', '..', 'skip_download_model')): + if not os.path.exists(ultralytics_bbox_path): + os.makedirs(ultralytics_bbox_path) + + if not os.path.exists(ultralytics_segm_path): + os.makedirs(ultralytics_segm_path) + + if not os.path.exists(os.path.join(ultralytics_bbox_path, "face_yolov8m.pt")): + download_url("https://huggingface.co/Bingsu/adetailer/resolve/main/face_yolov8m.pt", + ultralytics_bbox_path) + + if not os.path.exists(os.path.join(ultralytics_bbox_path, "hand_yolov8s.pt")): + download_url("https://huggingface.co/Bingsu/adetailer/resolve/main/hand_yolov8s.pt", + ultralytics_bbox_path) + + if not os.path.exists(os.path.join(ultralytics_segm_path, "person_yolov8m-seg.pt")): + download_url("https://huggingface.co/Bingsu/adetailer/resolve/main/person_yolov8m-seg.pt", + ultralytics_segm_path) diff --git a/ComfyUI-Impact-Pack/impact_subpack/requirements.txt b/ComfyUI-Impact-Pack/impact_subpack/requirements.txt new file mode 100644 index 0000000000000000000000000000000000000000..142db6c56283e520671827a52af363ff23b2c153 --- /dev/null +++ b/ComfyUI-Impact-Pack/impact_subpack/requirements.txt @@ -0,0 +1,2 @@ +matplotlib +ultralytics!=8.2.87 diff --git a/ComfyUI-Impact-Pack/install-manual.py b/ComfyUI-Impact-Pack/install-manual.py new file mode 100644 index 0000000000000000000000000000000000000000..d37226a68b454db6a981ec65f85f5c6e29012c80 --- /dev/null +++ b/ComfyUI-Impact-Pack/install-manual.py @@ -0,0 +1,183 @@ +import functools +import os +import re +import shutil +import sys +import subprocess +import threading +import locale +import traceback +from typing import Set + + +if sys.argv[0] == 'install.py': + sys.path.append('.') # for portable version + + +impact_path = os.path.join(os.path.dirname(__file__), "modules") +subpack_path = os.path.join(os.path.dirname(__file__), "impact_subpack") +subpack_repo = "https://github.com/ltdrdata/ComfyUI-Impact-Subpack" + + +comfy_path = os.environ.get('COMFYUI_PATH') +if comfy_path is None: + print(f"\n[bold yellow]WARN: The `COMFYUI_PATH` environment variable is not set. Assuming `{os.path.dirname(__file__)}/../../` as the ComfyUI path.[/bold yellow]", file=sys.stderr) + comfy_path = os.path.abspath(os.path.join(os.path.dirname(__file__), '..', '..')) + +model_path = os.environ.get('COMFYUI_MODEL_PATH') +if model_path is None: + try: + import folder_paths + model_path = folder_paths.models_dir + except: + pass + + if model_path is None: + model_path = os.path.abspath(os.path.join(comfy_path, 'models')) + print(f"\n[bold yellow]WARN: The `COMFYUI_MODEL_PATH` environment variable is not set. Assuming `{model_path}` as the ComfyUI path.[/bold yellow]", file=sys.stderr) + + +sys.path.append(impact_path) +sys.path.append(comfy_path) + + +# --- +def handle_stream(stream, is_stdout): + stream.reconfigure(encoding=locale.getpreferredencoding(), errors='replace') + + for msg in stream: + if is_stdout: + print(msg, end="", file=sys.stdout) + else: + print(msg, end="", file=sys.stderr) + + +def process_wrap(cmd_str, cwd=None, handler=None, env=None): + print(f"[Impact Pack] EXECUTE: {cmd_str} in '{cwd}'") + process = subprocess.Popen(cmd_str, cwd=cwd, stdout=subprocess.PIPE, stderr=subprocess.PIPE, env=env, text=True, bufsize=1) + + if handler is None: + handler = handle_stream + + stdout_thread = threading.Thread(target=handler, args=(process.stdout, True)) + stderr_thread = threading.Thread(target=handler, args=(process.stderr, False)) + + stdout_thread.start() + stderr_thread.start() + + stdout_thread.join() + stderr_thread.join() + + return process.wait() + + +@functools.cache +def get_installed_packages() -> Set[str]: + try: + result = subprocess.check_output([sys.executable, '-m', 'pip', 'list'], universal_newlines=True) + pip_list = set([line.split()[0].lower() for line in result.split('\n') if line.strip()]) + return pip_list + except subprocess.CalledProcessError as e: + raise Exception(f"[ComfyUI-Impact-Pack] Failed to retrieve the information of installed pip packages.") + + +def is_package_installed(name: str) -> bool: + name = name.strip() + pattern = r'([^<>!=]+)([<>!=]=?)' + match = re.search(pattern, name) + + if match: + name = match.group(1) + + result = name.lower() in get_installed_packages() + return result + + +def is_requirements_installed(file_path: str) -> bool: + print(f"Requirements file: {file_path}") + if os.path.exists(file_path): + with open(file_path, "r") as file: + lines = file.readlines() + + for line in lines: + if not is_package_installed(line): + return False + +pip_install = [sys.executable, "-m", "pip", "install", "-U"] + +# --- + + +try: + import platform + import folder_paths + from torchvision.datasets.utils import download_url + import impact.config + + print("### ComfyUI-Impact-Pack: Check dependencies") + def ensure_subpack(): + import git + if os.path.exists(subpack_path): + try: + repo = git.Repo(subpack_path) + repo.remotes.origin.pull() + except: + traceback.print_exc() + if platform.system() == 'Windows': + print(f"[ComfyUI-Impact-Pack] Please turn off ComfyUI and remove '{subpack_path}' and restart ComfyUI.") + else: + shutil.rmtree(subpack_path) + git.Repo.clone_from(subpack_repo, subpack_path) + else: + git.Repo.clone_from(subpack_repo, subpack_path) + + + def install(): + subpack_install_script = os.path.join(subpack_path, "install.py") + + print(f"### ComfyUI-Impact-Pack: Updating subpack") + ensure_subpack() # The installation of the subpack must take place before ensure_pip. cv2 triggers a permission error. + + new_env = os.environ.copy() + new_env["COMFYUI_PATH"] = comfy_path + new_env["COMFYUI_MODEL_PATH"] = model_path + + if os.path.exists(subpack_install_script): + if not is_requirements_installed(os.path.join(subpack_path, 'requirements.txt')): + process_wrap(pip_install + ['-r', 'requirements.txt'], cwd=subpack_path) + + process_wrap([sys.executable, 'install.py'], cwd=subpack_path, env=new_env) + else: + print(f"### ComfyUI-Impact-Pack: (Install Failed) Subpack\nFile not found: `{subpack_install_script}`") + + # Download model + print("### ComfyUI-Impact-Pack: Check basic models") + sam_path = os.path.join(model_path, "sams") + onnx_path = os.path.join(model_path, "onnx") + + if not os.path.exists(os.path.join(os.path.dirname(__file__), '..', 'skip_download_model')): + if not impact.config.get_config()['mmdet_skip']: + bbox_path = os.path.join(model_path, "mmdets", "bbox") + if not os.path.exists(bbox_path): + os.makedirs(bbox_path) + + if not os.path.exists(os.path.join(bbox_path, "mmdet_anime-face_yolov3.pth")): + download_url("https://huggingface.co/dustysys/ddetailer/resolve/main/mmdet/bbox/mmdet_anime-face_yolov3.pth", bbox_path) + + if not os.path.exists(os.path.join(bbox_path, "mmdet_anime-face_yolov3.py")): + download_url("https://raw.githubusercontent.com/Bing-su/dddetailer/master/config/mmdet_anime-face_yolov3.py", bbox_path) + + if not os.path.exists(os.path.join(sam_path, "sam_vit_b_01ec64.pth")): + download_url("https://dl.fbaipublicfiles.com/segment_anything/sam_vit_b_01ec64.pth", sam_path) + + if not os.path.exists(onnx_path): + print(f"### ComfyUI-Impact-Pack: onnx model directory created ({onnx_path})") + os.mkdir(onnx_path) + + impact.config.write_config() + + install() + +except Exception as e: + print("[ERROR] ComfyUI-Impact-Pack: Dependency installation has failed. Please install manually.") + traceback.print_exc() diff --git a/ComfyUI-Impact-Pack/install.py b/ComfyUI-Impact-Pack/install.py new file mode 100644 index 0000000000000000000000000000000000000000..d46db2d11d1af9d0c19ef7d1b43d82b5a3480220 --- /dev/null +++ b/ComfyUI-Impact-Pack/install.py @@ -0,0 +1,141 @@ +import os +import shutil +import sys +import subprocess +import threading +import locale +import traceback + + +if sys.argv[0] == 'install.py': + sys.path.append('.') # for portable version + + +impact_path = os.path.join(os.path.dirname(__file__), "modules") +subpack_path = os.path.join(os.path.dirname(__file__), "impact_subpack") +subpack_repo = "https://github.com/ltdrdata/ComfyUI-Impact-Subpack" + + +comfy_path = os.environ.get('COMFYUI_PATH') +if comfy_path is None: + print(f"\n[bold yellow]WARN: The `COMFYUI_PATH` environment variable is not set. Assuming `{os.path.dirname(__file__)}/../../` as the ComfyUI path.[/bold yellow]", file=sys.stderr) + comfy_path = os.path.abspath(os.path.join(os.path.dirname(__file__), '..', '..')) + +model_path = os.environ.get('COMFYUI_MODEL_PATH') +if model_path is None: + try: + import folder_paths + model_path = folder_paths.models_dir + except: + pass + + if model_path is None: + model_path = os.path.abspath(os.path.join(comfy_path, 'models')) + print(f"\n[bold yellow]WARN: The `COMFYUI_MODEL_PATH` environment variable is not set. Assuming `{model_path}` as the ComfyUI path.[/bold yellow]", file=sys.stderr) + + +sys.path.append(impact_path) +sys.path.append(comfy_path) + + +# --- +def handle_stream(stream, is_stdout): + stream.reconfigure(encoding=locale.getpreferredencoding(), errors='replace') + + for msg in stream: + if is_stdout: + print(msg, end="", file=sys.stdout) + else: + print(msg, end="", file=sys.stderr) + + +def process_wrap(cmd_str, cwd=None, handler=None, env=None): + print(f"[Impact Pack] EXECUTE: {cmd_str} in '{cwd}'") + process = subprocess.Popen(cmd_str, cwd=cwd, stdout=subprocess.PIPE, stderr=subprocess.PIPE, env=env, text=True, bufsize=1) + + if handler is None: + handler = handle_stream + + stdout_thread = threading.Thread(target=handler, args=(process.stdout, True)) + stderr_thread = threading.Thread(target=handler, args=(process.stderr, False)) + + stdout_thread.start() + stderr_thread.start() + + stdout_thread.join() + stderr_thread.join() + + return process.wait() +# --- + + +try: + import platform + import folder_paths + from torchvision.datasets.utils import download_url + import impact.config + + print("### ComfyUI-Impact-Pack: Check dependencies") + def ensure_subpack(): + import git + if os.path.exists(subpack_path): + try: + repo = git.Repo(subpack_path) + repo.remotes.origin.pull() + except: + traceback.print_exc() + if platform.system() == 'Windows': + print(f"[ComfyUI-Impact-Pack] Please turn off ComfyUI and remove '{subpack_path}' and restart ComfyUI.") + else: + shutil.rmtree(subpack_path) + git.Repo.clone_from(subpack_repo, subpack_path) + else: + git.Repo.clone_from(subpack_repo, subpack_path) + + + def install(): + subpack_install_script = os.path.join(subpack_path, "install.py") + + print(f"### ComfyUI-Impact-Pack: Updating subpack") + ensure_subpack() # The installation of the subpack must take place before ensure_pip. cv2 triggers a permission error. + + new_env = os.environ.copy() + new_env["COMFYUI_PATH"] = comfy_path + new_env["COMFYUI_MODEL_PATH"] = model_path + + if os.path.exists(subpack_install_script): + process_wrap([sys.executable, 'install.py'], cwd=subpack_path, env=new_env) + else: + print(f"### ComfyUI-Impact-Pack: (Install Failed) Subpack\nFile not found: `{subpack_install_script}`") + + # Download model + print("### ComfyUI-Impact-Pack: Check basic models") + sam_path = os.path.join(model_path, "sams") + onnx_path = os.path.join(model_path, "onnx") + + if not os.path.exists(os.path.join(os.path.dirname(__file__), '..', 'skip_download_model')): + if not impact.config.get_config()['mmdet_skip']: + bbox_path = os.path.join(model_path, "mmdets", "bbox") + if not os.path.exists(bbox_path): + os.makedirs(bbox_path) + + if not os.path.exists(os.path.join(bbox_path, "mmdet_anime-face_yolov3.pth")): + download_url("https://huggingface.co/dustysys/ddetailer/resolve/main/mmdet/bbox/mmdet_anime-face_yolov3.pth", bbox_path) + + if not os.path.exists(os.path.join(bbox_path, "mmdet_anime-face_yolov3.py")): + download_url("https://raw.githubusercontent.com/Bing-su/dddetailer/master/config/mmdet_anime-face_yolov3.py", bbox_path) + + if not os.path.exists(os.path.join(sam_path, "sam_vit_b_01ec64.pth")): + download_url("https://dl.fbaipublicfiles.com/segment_anything/sam_vit_b_01ec64.pth", sam_path) + + if not os.path.exists(onnx_path): + print(f"### ComfyUI-Impact-Pack: onnx model directory created ({onnx_path})") + os.mkdir(onnx_path) + + impact.config.write_config() + + install() + +except Exception as e: + print("[ERROR] ComfyUI-Impact-Pack: Dependency installation has failed. Please install manually.") + traceback.print_exc() diff --git a/ComfyUI-Impact-Pack/js/comboBoolMigration.js b/ComfyUI-Impact-Pack/js/comboBoolMigration.js new file mode 100644 index 0000000000000000000000000000000000000000..86982f134de4b3962b5e8287920964e33eb84f8d --- /dev/null +++ b/ComfyUI-Impact-Pack/js/comboBoolMigration.js @@ -0,0 +1,35 @@ +import { ComfyApp, app } from "../../scripts/app.js"; + +let conflict_check = undefined; + +app.registerExtension({ + name: "Comfy.impact.comboBoolMigration", + + nodeCreated(node, app) { + for(let i in node.widgets) { + let widget = node.widgets[i]; + + if(conflict_check == undefined) { + conflict_check = !!app.extensions.find((ext) => ext.name === "Comfy.comboBoolMigration"); + } + + if(conflict_check) + return; + + if(widget.type == "toggle") { + let value = widget.value; + + var v = Object.getOwnPropertyDescriptor(widget, 'value'); + if(!v) { + Object.defineProperty(widget, "value", { + set: (value) => { + delete widget.value; + widget.value = value == true || value == widget.options.on; + }, + get: () => { return value; } + }); + } + } + } + } +}); diff --git a/ComfyUI-Impact-Pack/js/common.js b/ComfyUI-Impact-Pack/js/common.js new file mode 100644 index 0000000000000000000000000000000000000000..aa5fbd5b7d96a8b60b058522a47bc9df5d7163e0 --- /dev/null +++ b/ComfyUI-Impact-Pack/js/common.js @@ -0,0 +1,95 @@ +import { api } from "../../scripts/api.js"; +import { app } from "../../scripts/app.js"; + +let original_show = app.ui.dialog.show; + +function dialog_show_wrapper(html) { + if (typeof html === "string") { + if(html.includes("IMPACT-PACK-SIGNAL: STOP CONTROL BRIDGE")) { + return; + } + + this.textElement.innerHTML = html; + } else { + this.textElement.replaceChildren(html); + } + this.element.style.display = "flex"; +} + +app.ui.dialog.show = dialog_show_wrapper; + + +function nodeFeedbackHandler(event) { + let nodes = app.graph._nodes_by_id; + let node = nodes[event.detail.node_id]; + if(node) { + const w = node.widgets.find((w) => event.detail.widget_name === w.name); + if(w) { + w.value = event.detail.value; + } + } +} + +api.addEventListener("impact-node-feedback", nodeFeedbackHandler); + + +function setMuteState(event) { + let nodes = app.graph._nodes_by_id; + let node = nodes[event.detail.node_id]; + if(node) { + if(event.detail.is_active) + node.mode = 0; + else + node.mode = 2; + } +} + +api.addEventListener("impact-node-mute-state", setMuteState); + + +async function bridgeContinue(event) { + let nodes = app.graph._nodes_by_id; + let node = nodes[event.detail.node_id]; + if(node) { + const mutes = new Set(event.detail.mutes); + const actives = new Set(event.detail.actives); + const bypasses = new Set(event.detail.bypasses); + + for(let i in app.graph._nodes_by_id) { + let this_node = app.graph._nodes_by_id[i]; + if(mutes.has(i)) { + this_node.mode = 2; + } + else if(actives.has(i)) { + this_node.mode = 0; + } + else if(bypasses.has(i)) { + this_node.mode = 4; + } + } + + await app.queuePrompt(0, 1); + } +} + +api.addEventListener("impact-bridge-continue", bridgeContinue); + + +function addQueue(event) { + app.queuePrompt(0, 1); +} + +api.addEventListener("impact-add-queue", addQueue); + + +function refreshPreview(event) { + let node_id = event.detail.node_id; + let item = event.detail.item; + let img = new Image(); + img.src = `/view?filename=${item.filename}&subfolder=${item.subfolder}&type=${item.type}&no-cache=${Date.now()}`; + let node = app.graph._nodes_by_id[node_id]; + if(node) + node.imgs = [img]; +} + +api.addEventListener("impact-preview", refreshPreview); diff --git a/ComfyUI-Impact-Pack/js/impact-image-util.js b/ComfyUI-Impact-Pack/js/impact-image-util.js new file mode 100644 index 0000000000000000000000000000000000000000..678d60ad5766b80b7deeb538e0c63bc9e6384d19 --- /dev/null +++ b/ComfyUI-Impact-Pack/js/impact-image-util.js @@ -0,0 +1,229 @@ +import { ComfyApp, app } from "../../scripts/app.js"; +import { api } from "../../scripts/api.js"; + +function load_image(str) { + let base64String = canvas.toDataURL('image/png'); + let img = new Image(); + img.src = base64String; +} + +function getFileItem(baseType, path) { + try { + let pathType = baseType; + + if (path.endsWith("[output]")) { + pathType = "output"; + path = path.slice(0, -9); + } else if (path.endsWith("[input]")) { + pathType = "input"; + path = path.slice(0, -8); + } else if (path.endsWith("[temp]")) { + pathType = "temp"; + path = path.slice(0, -7); + } + + const subfolder = path.substring(0, path.lastIndexOf('/')); + const filename = path.substring(path.lastIndexOf('/') + 1); + + return { + filename: filename, + subfolder: subfolder, + type: pathType + }; + } + catch(exception) { + return null; + } +} + +async function loadImageFromUrl(image, node_id, v, need_to_load) { + let item = getFileItem('temp', v); + + if(item) { + let params = `?node_id=${node_id}&filename=${item.filename}&type=${item.type}&subfolder=${item.subfolder}`; + + let res = await api.fetchApi('/impact/set/pb_id_image'+params, { cache: "no-store" }); + if(res.status == 200) { + let pb_id = await res.text(); + if(need_to_load) {; + image.src = api.apiURL(`/view?filename=${item.filename}&type=${item.type}&subfolder=${item.subfolder}`); + } + return pb_id; + } + else { + return `$${node_id}-0`; + } + } + else { + return `$${node_id}-0`; + } +} + +async function loadImageFromId(image, v) { + let res = await api.fetchApi('/impact/get/pb_id_image?id='+v, { cache: "no-store" }); + if(res.status == 200) { + let item = await res.json(); + image.src = api.apiURL(`/view?filename=${item.filename}&type=${item.type}&subfolder=${item.subfolder}`); + return true; + } + + return false; +} + +app.registerExtension({ + name: "Comfy.Impact.img", + + nodeCreated(node, app) { + if(node.comfyClass == "PreviewBridge" || node.comfyClass == "PreviewBridgeLatent") { + let w = node.widgets.find(obj => obj.name === 'image'); + node._imgs = [new Image()]; + node.imageIndex = 0; + + Object.defineProperty(w, 'value', { + async set(v) { + if(w._lock) + return; + + const stackTrace = new Error().stack; + if(stackTrace.includes('presetText.js')) + return; + + var image = new Image(); + if(v && v.constructor == String && v.startsWith('$')) { + // from node feedback + let need_to_load = node._imgs[0].src == ''; + if(await loadImageFromId(image, v, need_to_load)) { + w._value = v; + if(node._imgs[0].src == '') { + node._imgs = [image]; + } + } + else { + w._value = `$${node.id}-0`; + } + } + else { + // from clipspace + w._lock = true; + w._value = await loadImageFromUrl(image, node.id, v, false); + w._lock = false; + } + }, + get() { + if(w._value == undefined) { + w._value = `$${node.id}-0`; + } + return w._value; + } + }); + + Object.defineProperty(node, 'imgs', { + set(v) { + const stackTrace = new Error().stack; + if(v && v.length == 0) + return; + else if(stackTrace.includes('pasteFromClipspace')) { + let sp = new URLSearchParams(v[0].src.split("?")[1]); + let str = ""; + if(sp.get('subfolder')) { + str += sp.get('subfolder') + '/'; + } + str += `${sp.get("filename")} [${sp.get("type")}]`; + + w.value = str; + } + + node._imgs = v; + }, + get() { + return node._imgs; + } + }); + } + + if(node.comfyClass == "ImageReceiver") { + let path_widget = node.widgets.find(obj => obj.name === 'image'); + let w = node.widgets.find(obj => obj.name === 'image_data'); + let stw_widget = node.widgets.find(obj => obj.name === 'save_to_workflow'); + w._value = ""; + + Object.defineProperty(w, 'value', { + set(v) { + if(v != '[IMAGE DATA]') + w._value = v; + }, + get() { + const stackTrace = new Error().stack; + if(!stackTrace.includes('draw') && !stackTrace.includes('graphToPrompt') && stackTrace.includes('app.js')) { + return "[IMAGE DATA]"; + } + else { + if(stw_widget.value) + return w._value; + else + return ""; + } + } + }); + + let set_img_act = (v) => { + node._img = v; + var canvas = document.createElement('canvas'); + canvas.width = v[0].width; + canvas.height = v[0].height; + + var context = canvas.getContext('2d'); + context.drawImage(v[0], 0, 0, v[0].width, v[0].height); + + var base64Image = canvas.toDataURL('image/png'); + w.value = base64Image; + }; + + Object.defineProperty(node, 'imgs', { + set(v) { + if (v && !v[0].complete) { + let orig_onload = v[0].onload; + v[0].onload = function(v2) { + if(orig_onload) + orig_onload(); + set_img_act(v); + }; + } + else { + set_img_act(v); + } + }, + get() { + if(this._img == undefined && w.value != '') { + this._img = [new Image()]; + if(stw_widget.value && w.value != '[IMAGE DATA]') + this._img[0].src = w.value; + } + else if(this._img == undefined && path_widget.value) { + let image = new Image(); + image.src = path_widget.value; + + try { + let item = getFileItem('temp', path_widget.value); + let params = `?filename=${item.filename}&type=${item.type}&subfolder=${item.subfolder}`; + + let res = api.fetchApi('/view/validate'+params, { cache: "no-store" }).then(response => response); + if(res.status == 200) { + image.src = api.apiURL('/view'+params); + } + + this._img = [new Image()]; // placeholder + image.onload = function(v) { + set_img_act([image]); + }; + } + catch { + + } + } + return this._img; + } + }); + } + } +}) diff --git a/ComfyUI-Impact-Pack/js/impact-pack.js b/ComfyUI-Impact-Pack/js/impact-pack.js new file mode 100644 index 0000000000000000000000000000000000000000..ac0f2f64b269b8ebbb897e1e76ef48173e764a06 --- /dev/null +++ b/ComfyUI-Impact-Pack/js/impact-pack.js @@ -0,0 +1,829 @@ +import { ComfyApp, app } from "../../scripts/app.js"; +import { ComfyDialog, $el } from "../../scripts/ui.js"; +import { api } from "../../scripts/api.js"; + +let wildcards_list = []; +async function load_wildcards() { + let res = await api.fetchApi('/impact/wildcards/list'); + let data = await res.json(); + wildcards_list = data.data; +} + +load_wildcards(); + +export function get_wildcards_list() { + return wildcards_list; +} + +// temporary implementation (copying from https://github.com/pythongosssss/ComfyUI-WD14-Tagger) +// I think this should be included into master!! +class ImpactProgressBadge { + constructor() { + if (!window.__progress_badge__) { + window.__progress_badge__ = Symbol("__impact_progress_badge__"); + } + this.symbol = window.__progress_badge__; + } + + getState(node) { + return node[this.symbol] || {}; + } + + setState(node, state) { + node[this.symbol] = state; + app.canvas.setDirty(true); + } + + addStatusHandler(nodeType) { + if (nodeType[this.symbol]?.statusTagHandler) { + return; + } + if (!nodeType[this.symbol]) { + nodeType[this.symbol] = {}; + } + nodeType[this.symbol] = { + statusTagHandler: true, + }; + + api.addEventListener("impact/update_status", ({ detail }) => { + let { node, progress, text } = detail; + const n = app.graph.getNodeById(+(node || app.runningNodeId)); + if (!n) return; + const state = this.getState(n); + state.status = Object.assign(state.status || {}, { progress: text ? progress : null, text: text || null }); + this.setState(n, state); + }); + + const self = this; + const onDrawForeground = nodeType.prototype.onDrawForeground; + nodeType.prototype.onDrawForeground = function (ctx) { + const r = onDrawForeground?.apply?.(this, arguments); + const state = self.getState(this); + if (!state?.status?.text) { + return r; + } + + const { fgColor, bgColor, text, progress, progressColor } = { ...state.status }; + + ctx.save(); + ctx.font = "12px sans-serif"; + const sz = ctx.measureText(text); + ctx.fillStyle = bgColor || "dodgerblue"; + ctx.beginPath(); + ctx.roundRect(0, -LiteGraph.NODE_TITLE_HEIGHT - 20, sz.width + 12, 20, 5); + ctx.fill(); + + if (progress) { + ctx.fillStyle = progressColor || "green"; + ctx.beginPath(); + ctx.roundRect(0, -LiteGraph.NODE_TITLE_HEIGHT - 20, (sz.width + 12) * progress, 20, 5); + ctx.fill(); + } + + ctx.fillStyle = fgColor || "#fff"; + ctx.fillText(text, 6, -LiteGraph.NODE_TITLE_HEIGHT - 6); + ctx.restore(); + return r; + }; + } +} + +const input_tracking = {}; +const input_dirty = {}; +const output_tracking = {}; + +function progressExecuteHandler(event) { + if(event.detail.output.aux){ + const id = event.detail.node; + if(input_tracking.hasOwnProperty(id)) { + if(input_tracking.hasOwnProperty(id) && input_tracking[id][0] != event.detail.output.aux[0]) { + input_dirty[id] = true; + } + else{ + + } + } + + input_tracking[id] = event.detail.output.aux; + } +} + +function imgSendHandler(event) { + if(event.detail.images.length > 0){ + let data = event.detail.images[0]; + let filename = `${data.filename} [${data.type}]`; + + let nodes = app.graph._nodes; + for(let i in nodes) { + if(nodes[i].type == 'ImageReceiver') { + let is_linked = false; + + if(nodes[i].widgets[1].type == 'converted-widget') { + for(let j in nodes[i].inputs) { + let input = nodes[i].inputs[j]; + if(input.name === 'link_id') { + if(input.link) { + let src_node = app.graph._nodes_by_id[app.graph.links[input.link].origin_id]; + if(src_node.type == 'ImpactInt' || src_node.type == 'PrimitiveNode') { + is_linked = true; + } + } + break; + } + } + } + else if(nodes[i].widgets[1].value == event.detail.link_id) { + is_linked = true; + } + + if(is_linked) { + if(data.subfolder) + nodes[i].widgets[0].value = `${data.subfolder}/${data.filename} [${data.type}]`; + else + nodes[i].widgets[0].value = `${data.filename} [${data.type}]`; + + let img = new Image(); + img.onload = (event) => { + nodes[i].imgs = [img]; + nodes[i].size[1] = Math.max(200, nodes[i].size[1]); + app.canvas.setDirty(true); + }; + img.src = `/view?filename=${data.filename}&type=${data.type}&subfolder=${data.subfolder}`+app.getPreviewFormatParam(); + } + } + } + } +} + + +function latentSendHandler(event) { + if(event.detail.images.length > 0){ + let data = event.detail.images[0]; + let filename = `${data.filename} [${data.type}]`; + + let nodes = app.graph._nodes; + for(let i in nodes) { + if(nodes[i].type == 'LatentReceiver') { + if(nodes[i].widgets[1].value == event.detail.link_id) { + if(data.subfolder) + nodes[i].widgets[0].value = `${data.subfolder}/${data.filename} [${data.type}]`; + else + nodes[i].widgets[0].value = `${data.filename} [${data.type}]`; + + let img = new Image(); + img.src = `/view?filename=${data.filename}&type=${data.type}&subfolder=${data.subfolder}`+app.getPreviewFormatParam(); + nodes[i].imgs = [img]; + nodes[i].size[1] = Math.max(200, nodes[i].size[1]); + } + } + } + } +} + + +function valueSendHandler(event) { + let nodes = app.graph._nodes; + for(let i in nodes) { + if(nodes[i].type == 'ImpactValueReceiver') { + if(nodes[i].widgets[2].value == event.detail.link_id) { + nodes[i].widgets[1].value = event.detail.value; + + let typ = typeof event.detail.value; + if(typ == 'string') { + nodes[i].widgets[0].value = "STRING"; + } + else if(typ == "boolean") { + nodes[i].widgets[0].value = "BOOLEAN"; + } + else if(typ != "number") { + nodes[i].widgets[0].value = typeof event.detail.value; + } + else if(Number.isInteger(event.detail.value)) { + nodes[i].widgets[0].value = "INT"; + } + else { + nodes[i].widgets[0].value = "FLOAT"; + } + } + } + } +} + + +const impactProgressBadge = new ImpactProgressBadge(); + +api.addEventListener("stop-iteration", () => { + document.getElementById("autoQueueCheckbox").checked = false; +}); +api.addEventListener("value-send", valueSendHandler); +api.addEventListener("img-send", imgSendHandler); +api.addEventListener("latent-send", latentSendHandler); +api.addEventListener("executed", progressExecuteHandler); + +app.registerExtension({ + name: "Comfy.Impack", + loadedGraphNode(node, app) { + if (node.comfyClass == "MaskPainter") { + input_dirty[node.id + ""] = true; + } + }, + + async beforeRegisterNodeDef(nodeType, nodeData, app) { + if (nodeData.name == "IterativeLatentUpscale" || nodeData.name == "IterativeImageUpscale" + || nodeData.name == "RegionalSampler"|| nodeData.name == "RegionalSamplerAdvanced") { + impactProgressBadge.addStatusHandler(nodeType); + } + + if(nodeData.name == "ImpactControlBridge") { + const onConnectionsChange = nodeType.prototype.onConnectionsChange; + nodeType.prototype.onConnectionsChange = function (type, index, connected, link_info) { + if(index != 0 || !link_info || this.inputs[0].type != '*') + return; + + // assign type + let slot_type = '*'; + + if(type == 2) { + slot_type = link_info.type; + } + else { + const node = app.graph.getNodeById(link_info.origin_id); + slot_type = node.outputs[link_info.origin_slot].type; + } + + this.inputs[0].type = slot_type; + this.outputs[0].type = slot_type; + this.outputs[0].label = slot_type; + } + } + + if(nodeData.name == "ImpactConditionalBranch" || nodeData.name == "ImpactConditionalBranchSelMode") { + const onConnectionsChange = nodeType.prototype.onConnectionsChange; + nodeType.prototype.onConnectionsChange = function (type, index, connected, link_info) { + if(!link_info || this.inputs[0].type != '*') + return; + + if(index >= 2) + return; + + // assign type + let slot_type = '*'; + + if(type == 2) { + slot_type = link_info.type; + } + else { + const node = app.graph.getNodeById(link_info.origin_id); + slot_type = node.outputs[link_info.origin_slot].type; + } + + this.inputs[0].type = slot_type; + this.inputs[1].type = slot_type; + this.outputs[0].type = slot_type; + this.outputs[0].label = slot_type; + } + } + + if(nodeData.name == "ImpactCompare") { + const onConnectionsChange = nodeType.prototype.onConnectionsChange; + nodeType.prototype.onConnectionsChange = function (type, index, connected, link_info) { + if(!link_info || this.inputs[0].type != '*' || type == 2) + return; + + // assign type + const node = app.graph.getNodeById(link_info.origin_id); + let slot_type = node.outputs[link_info.origin_slot].type; + + this.inputs[0].type = slot_type; + this.inputs[1].type = slot_type; + } + } + + if(nodeData.name === 'ImpactInversedSwitch') { + nodeData.output = ['*']; + nodeData.output_is_list = [false]; + nodeData.output_name = ['output1']; + + const onConnectionsChange = nodeType.prototype.onConnectionsChange; + nodeType.prototype.onConnectionsChange = function (type, index, connected, link_info) { + if(!link_info) + return; + + if(type == 2) { + // connect output + if(connected){ + if(app.graph._nodes_by_id[link_info.target_id].type == 'Reroute') { + app.graph._nodes_by_id[link_info.target_id].disconnectInput(link_info.target_slot); + } + + if(this.outputs[0].type == '*'){ + if(link_info.type == '*') { + app.graph._nodes_by_id[link_info.target_id].disconnectInput(link_info.target_slot); + } + else { + // propagate type + this.outputs[0].type = link_info.type; + this.outputs[0].name = link_info.type; + + for(let i in this.inputs) { + if(this.inputs[i].name != 'select') + this.inputs[i].type = link_info.type; + } + } + } + } + } + else { + if(app.graph._nodes_by_id[link_info.origin_id].type == 'Reroute') + this.disconnectInput(link_info.target_slot); + + // connect input + if(this.inputs[0].type == '*'){ + const node = app.graph.getNodeById(link_info.origin_id); + let origin_type = node.outputs[link_info.origin_slot].type; + + if(origin_type == '*') { + this.disconnectInput(link_info.target_slot); + return; + } + + for(let i in this.inputs) { + if(this.inputs[i].name != 'select') + this.inputs[i].type = origin_type; + } + + this.outputs[0].type = origin_type; + this.outputs[0].name = 'output1'; + } + + return; + } + + if (!connected && this.outputs.length > 1) { + const stackTrace = new Error().stack; + + if( + !stackTrace.includes('LGraphNode.prototype.connect') && // for touch device + !stackTrace.includes('LGraphNode.connect') && // for mouse device + !stackTrace.includes('loadGraphData')) { + if(this.outputs[link_info.origin_slot].links.length == 0) + this.removeOutput(link_info.origin_slot); + } + } + + let slot_i = 1; + for (let i = 0; i < this.outputs.length; i++) { + this.outputs[i].name = `output${slot_i}` + slot_i++; + } + + let last_slot = this.outputs[this.outputs.length - 1]; + if (last_slot.slot_index == link_info.origin_slot) { + this.addOutput(`output${slot_i}`, this.outputs[0].type); + } + + let select_slot = this.inputs.find(x => x.name == "select"); + if(this.widgets?.length) { + this.widgets[0].options.max = select_slot?this.outputs.length-1:this.outputs.length; + this.widgets[0].value = Math.min(this.widgets[0].value, this.widgets[0].options.max); + if(this.widgets[0].options.max > 0 && this.widgets[0].value == 0) + this.widgets[0].value = 1; + } + } + } + + if (nodeData.name === 'ImpactMakeImageList' || nodeData.name === 'ImpactMakeImageBatch' || + nodeData.name === 'ImpactMakeMaskList' || nodeData.name === 'ImpactMakeMaskBatch' || + nodeData.name === 'ImpactMakeAnyList' || nodeData.name === 'CombineRegionalPrompts' || + nodeData.name === 'ImpactCombineConditionings' || nodeData.name === 'ImpactConcatConditionings' || + nodeData.name === 'ImpactSEGSConcat' || + nodeData.name === 'ImpactSwitch' || nodeData.name === 'LatentSwitch' || nodeData.name == 'SEGSSwitch') { + var input_name = "input"; + + switch(nodeData.name) { + case 'ImpactMakeImageList': + case 'ImpactMakeImageBatch': + input_name = "image"; + break; + + case 'ImpactMakeMaskList': + case 'ImpactMakeMaskBatch': + input_name = "mask"; + break; + + case 'ImpactMakeAnyList': + input_name = "value"; + break; + + case 'ImpactSEGSConcat': + input_name = "segs"; + break; + + case 'CombineRegionalPrompts': + input_name = "regional_prompts"; + break; + + case 'ImpactCombineConditionings': + case 'ImpactConcatConditionings': + input_name = "conditioning"; + break; + + case 'LatentSwitch': + input_name = "input"; + break; + + case 'SEGSSwitch': + input_name = "input"; + break; + + case 'ImpactSwitch': + input_name = "input"; + } + + const onConnectionsChange = nodeType.prototype.onConnectionsChange; + nodeType.prototype.onConnectionsChange = function (type, index, connected, link_info) { + if(!link_info) + return; + + if(type == 2) { + // connect output + if(connected && index == 0){ + if(nodeData.name == 'ImpactSwitch' && app.graph._nodes_by_id[link_info.target_id]?.type == 'Reroute') { + app.graph._nodes_by_id[link_info.target_id].disconnectInput(link_info.target_slot); + } + + if(this.outputs[0].type == '*'){ + if(link_info.type == '*') { + app.graph._nodes_by_id[link_info.target_id].disconnectInput(link_info.target_slot); + } + else { + // propagate type + this.outputs[0].type = link_info.type; + this.outputs[0].label = link_info.type; + this.outputs[0].name = link_info.type; + + for(let i in this.inputs) { + let input_i = this.inputs[i]; + if(input_i.name != 'select' && input_i.name != 'sel_mode') + input_i.type = link_info.type; + } + } + } + } + + return; + } + else { + if(nodeData.name == 'ImpactSwitch' && app.graph._nodes_by_id[link_info.origin_id].type == 'Reroute') + this.disconnectInput(link_info.target_slot); + + // connect input + if(this.inputs[index].name == 'select' || this.inputs[index].name == 'sel_mode') + return; + + if(this.inputs[0].type == '*'){ + const node = app.graph.getNodeById(link_info.origin_id); + let origin_type = node.outputs[link_info.origin_slot]?.type; + if(link_info.target_slot == 0 && this.inputs.length > 1) { + origin_type = this.inputs[1].type; + node.connect(link_info.origin_slot, node.id, 'input1'); + } + + if(origin_type == '*') { + this.disconnectInput(link_info.target_slot); + return; + } + + for(let i in this.inputs) { + let input_i = this.inputs[i]; + if(input_i.name != 'select' && input_i.name != 'sel_mode') + input_i.type = origin_type; + } + + this.outputs[0].type = origin_type; + this.outputs[0].label = origin_type; + this.outputs[0].name = origin_type; + } + } + + let select_slot = this.inputs.find(x => x.name == "select"); + let mode_slot = this.inputs.find(x => x.name == "sel_mode"); + + let converted_count = 0; + converted_count += select_slot?1:0; + converted_count += mode_slot?1:0; + + if (!connected && (this.inputs.length > 1+converted_count)) { + const stackTrace = new Error().stack; + + if( + !stackTrace.includes('LGraphNode.prototype.connect') && // for touch device + !stackTrace.includes('LGraphNode.connect') && // for mouse device + !stackTrace.includes('loadGraphData') && + this.inputs[index].name != 'select') { + this.removeInput(index); + } + } + + let slot_i = 1; + for (let i = 0; i < this.inputs.length; i++) { + let input_i = this.inputs[i]; + if(input_i.name != 'select'&& input_i.name != 'sel_mode') { + input_i.name = `${input_name}${slot_i}` + slot_i++; + } + } + + let last_slot = this.inputs[this.inputs.length - 1]; + if ( + (last_slot.name == 'select' && last_slot.name != 'sel_mode' && this.inputs[this.inputs.length - 2].link != undefined) + || (last_slot.name != 'select' && last_slot.name != 'sel_mode' && last_slot.link != undefined)) { + this.addInput(`${input_name}${slot_i}`, this.outputs[0].type); + } + + if(this.widgets?.length) { + this.widgets[0].options.max = select_slot?this.inputs.length-1:this.inputs.length; + this.widgets[0].value = Math.min(this.widgets[0].value, this.widgets[0].options.max); + if(this.widgets[0].options.max > 0 && this.widgets[0].value == 0) + this.widgets[0].value = 1; + } + } + } + }, + + nodeCreated(node, app) { + if(node.comfyClass == "MaskPainter") { + node.addWidget("button", "Edit mask", null, () => { + ComfyApp.copyToClipspace(node); + ComfyApp.clipspace_return_node = node; + ComfyApp.open_maskeditor(); + }); + } + + switch(node.comfyClass) { + case "ToDetailerPipe": + case "ToDetailerPipeSDXL": + case "BasicPipeToDetailerPipe": + case "BasicPipeToDetailerPipeSDXL": + case "EditDetailerPipe": + case "FaceDetailer": + case "DetailerForEach": + case "DetailerForEachDebug": + case "DetailerForEachPipe": + case "DetailerForEachDebugPipe": + { + for(let i in node.widgets) { + let widget = node.widgets[i]; + if(widget.type === "customtext") { + widget.dynamicPrompts = false; + widget.inputEl.placeholder = "wildcard spec: if kept empty, this option will be ignored"; + widget.serializeValue = () => { + return node.widgets[i].value; + }; + } + } + } + break; + } + + if(node.comfyClass == "ImpactSEGSLabelFilter" || node.comfyClass == "SEGSLabelFilterDetailerHookProvider") { + Object.defineProperty(node.widgets[0], "value", { + set: (value) => { + const stackTrace = new Error().stack; + if(stackTrace.includes('inner_value_change')) { + if(node.widgets[1].value.trim() != "" && !node.widgets[1].value.trim().endsWith(",")) + node.widgets[1].value += ", " + + node.widgets[1].value += value; + node.widgets_values[1] = node.widgets[1].value; + } + + node._value = value; + }, + get: () => { + return node._value; + } + }); + } + + if(node.comfyClass == "UltralyticsDetectorProvider") { + let model_name_widget = node.widgets.find((w) => w.name === "model_name"); + let orig_draw = node.onDrawForeground; + node.onDrawForeground = function (ctx) { + const r = orig_draw?.apply?.(this, arguments); + + let is_seg = model_name_widget.value?.startsWith('segm/') || model_name_widget.value?.includes('-seg'); + if(!is_seg) { + var slot_pos = new Float32Array(2); + var pos = node.getConnectionPos(false, 1, slot_pos); + + pos[0] -= node.pos[0] - 10; + pos[1] -= node.pos[1]; + + ctx.beginPath(); + ctx.strokeStyle = "red"; + ctx.lineWidth = 4; + ctx.moveTo(pos[0] - 5, pos[1] - 5); + ctx.lineTo(pos[0] + 5, pos[1] + 5); + ctx.moveTo(pos[0] + 5, pos[1] - 5); + ctx.lineTo(pos[0] - 5, pos[1] + 5); + ctx.stroke(); + } + } + } + + if( + node.comfyClass == "ImpactWildcardEncode" || node.comfyClass == "ImpactWildcardProcessor" + || node.comfyClass == "ToDetailerPipe" || node.comfyClass == "ToDetailerPipeSDXL" + || node.comfyClass == "EditDetailerPipe" || node.comfyClass == "EditDetailerPipeSDXL" + || node.comfyClass == "BasicPipeToDetailerPipe" || node.comfyClass == "BasicPipeToDetailerPipeSDXL") { + node._value = "Select the LoRA to add to the text"; + node._wvalue = "Select the Wildcard to add to the text"; + + var tbox_id = 0; + var combo_id = 3; + var has_lora = true; + + switch(node.comfyClass){ + case "ImpactWildcardEncode": + tbox_id = 0; + combo_id = 3; + break; + + case "ImpactWildcardProcessor": + tbox_id = 0; + combo_id = 4; + has_lora = false; + break; + + case "ToDetailerPipe": + case "ToDetailerPipeSDXL": + case "EditDetailerPipe": + case "EditDetailerPipeSDXL": + case "BasicPipeToDetailerPipe": + case "BasicPipeToDetailerPipeSDXL": + tbox_id = 0; + combo_id = 1; + break; + } + + Object.defineProperty(node.widgets[combo_id+1], "value", { + set: (value) => { + const stackTrace = new Error().stack; + if(stackTrace.includes('inner_value_change')) { + if(value != "Select the Wildcard to add to the text") { + if(node.widgets[tbox_id].value != '') + node.widgets[tbox_id].value += ', ' + + node.widgets[tbox_id].value += value; + } + } + }, + get: () => { return "Select the Wildcard to add to the text"; } + }); + + Object.defineProperty(node.widgets[combo_id+1].options, "values", { + set: (x) => {}, + get: () => { + return wildcards_list; + } + }); + + if(has_lora) { + Object.defineProperty(node.widgets[combo_id], "value", { + set: (value) => { + const stackTrace = new Error().stack; + if(stackTrace.includes('inner_value_change')) { + if(value != "Select the LoRA to add to the text") { + let lora_name = value; + if (lora_name.endsWith('.safetensors')) { + lora_name = lora_name.slice(0, -12); + } + + node.widgets[tbox_id].value += ``; + if(node.widgets_values) { + node.widgets_values[tbox_id] = node.widgets[tbox_id].value; + } + } + } + + node._value = value; + }, + + get: () => { return "Select the LoRA to add to the text"; } + }); + } + + // Preventing validation errors from occurring in any situation. + if(has_lora) { + node.widgets[combo_id].serializeValue = () => { return "Select the LoRA to add to the text"; } + } + node.widgets[combo_id+1].serializeValue = () => { return "Select the Wildcard to add to the text"; } + } + + if(node.comfyClass == "ImpactWildcardProcessor" || node.comfyClass == "ImpactWildcardEncode") { + node.widgets[0].inputEl.placeholder = "Wildcard Prompt (User input)"; + node.widgets[1].inputEl.placeholder = "Populated Prompt (Will be generated automatically)"; + node.widgets[1].inputEl.disabled = true; + + const populated_text_widget = node.widgets.find((w) => w.name == 'populated_text'); + const mode_widget = node.widgets.find((w) => w.name == 'mode'); + + // mode combo + Object.defineProperty(mode_widget, "value", { + set: (value) => { + node._mode_value = value == true || value == "Populate"; + populated_text_widget.inputEl.disabled = value == true || value == "Populate"; + }, + get: () => { + if(node._mode_value != undefined) + return node._mode_value; + else + return true; + } + }); + } + + if (node.comfyClass == "MaskPainter") { + node.widgets[0].value = '#placeholder'; + + Object.defineProperty(node, "images", { + set: function(value) { + node._images = value; + }, + get: function() { + const id = node.id+""; + if(node.widgets[0].value != '#placeholder') { + var need_invalidate = false; + + if(input_dirty.hasOwnProperty(id) && input_dirty[id]) { + node.widgets[0].value = {...input_tracking[id][1]}; + input_dirty[id] = false; + need_invalidate = true + this._images = app.nodeOutputs[id].images; + } + + let filename = app.nodeOutputs[id]['aux'][1][0]['filename']; + let subfolder = app.nodeOutputs[id]['aux'][1][0]['subfolder']; + let type = app.nodeOutputs[id]['aux'][1][0]['type']; + + let item = + { + image_hash: app.nodeOutputs[id]['aux'][0], + forward_filename: app.nodeOutputs[id]['aux'][1][0]['filename'], + forward_subfolder: app.nodeOutputs[id]['aux'][1][0]['subfolder'], + forward_type: app.nodeOutputs[id]['aux'][1][0]['type'] + }; + + if(node._images) { + app.nodeOutputs[id].images = [{ + ...node._images[0], + ...item + }]; + + node.widgets[0].value = + { + ...node._images[0], + ...item + }; + } + else { + app.nodeOutputs[id].images = [{ + ...item + }]; + + node.widgets[0].value = + { + ...item + }; + } + + if(need_invalidate) { + Promise.all( + app.nodeOutputs[id].images.map((src) => { + return new Promise((r) => { + const img = new Image(); + img.onload = () => r(img); + img.onerror = () => r(null); + img.src = "/view?" + new URLSearchParams(src).toString(); + }); + }) + ).then((imgs) => { + this.imgs = imgs.filter(Boolean); + this.setSizeForImage?.(); + app.graph.setDirtyCanvas(true); + }); + + app.nodeOutputs[id].images[0] = { ...node.widgets[0].value }; + } + + return app.nodeOutputs[id].images; + } + else { + return node._images; + } + } + }); + } + } +}); diff --git a/ComfyUI-Impact-Pack/js/impact-sam-editor.js b/ComfyUI-Impact-Pack/js/impact-sam-editor.js new file mode 100644 index 0000000000000000000000000000000000000000..e7bf6f297fc67d41c3e25e6541cd0df98b51fdd1 --- /dev/null +++ b/ComfyUI-Impact-Pack/js/impact-sam-editor.js @@ -0,0 +1,637 @@ +import { app } from "../../scripts/app.js"; +import { api } from "../../scripts/api.js"; +import { ComfyDialog, $el } from "../../scripts/ui.js"; +import { ComfyApp } from "../../scripts/app.js"; +import { ClipspaceDialog } from "../../extensions/core/clipspace.js"; + +function addMenuHandler(nodeType, cb) { + const getOpts = nodeType.prototype.getExtraMenuOptions; + nodeType.prototype.getExtraMenuOptions = function () { + const r = getOpts.apply(this, arguments); + cb.apply(this, arguments); + return r; + }; +} + +// Helper function to convert a data URL to a Blob object +function dataURLToBlob(dataURL) { + const parts = dataURL.split(';base64,'); + const contentType = parts[0].split(':')[1]; + const byteString = atob(parts[1]); + const arrayBuffer = new ArrayBuffer(byteString.length); + const uint8Array = new Uint8Array(arrayBuffer); + for (let i = 0; i < byteString.length; i++) { + uint8Array[i] = byteString.charCodeAt(i); + } + return new Blob([arrayBuffer], { type: contentType }); +} + +function loadedImageToBlob(image) { + const canvas = document.createElement('canvas'); + + canvas.width = image.width; + canvas.height = image.height; + + const ctx = canvas.getContext('2d'); + + ctx.drawImage(image, 0, 0); + + const dataURL = canvas.toDataURL('image/png', 1); + const blob = dataURLToBlob(dataURL); + + return blob; +} + +async function uploadMask(filepath, formData) { + await api.fetchApi('/upload/mask', { + method: 'POST', + body: formData + }).then(response => {}).catch(error => { + console.error('Error:', error); + }); + + ComfyApp.clipspace.imgs[ComfyApp.clipspace['selectedIndex']] = new Image(); + ComfyApp.clipspace.imgs[ComfyApp.clipspace['selectedIndex']].src = `view?filename=${filepath.filename}&type=${filepath.type}`; + + if(ComfyApp.clipspace.images) + ComfyApp.clipspace.images[ComfyApp.clipspace['selectedIndex']] = filepath; + + ClipspaceDialog.invalidatePreview(); +} + +class ImpactSamEditorDialog extends ComfyDialog { + static instance = null; + + static getInstance() { + if(!ImpactSamEditorDialog.instance) { + ImpactSamEditorDialog.instance = new ImpactSamEditorDialog(); + } + + return ImpactSamEditorDialog.instance; + } + + constructor() { + super(); + this.element = $el("div.comfy-modal", { parent: document.body }, + [ $el("div.comfy-modal-content", + [...this.createButtons()]), + ]); + } + + createButtons() { + return []; + } + + createButton(name, callback) { + var button = document.createElement("button"); + button.innerText = name; + button.addEventListener("click", callback); + return button; + } + + createLeftButton(name, callback) { + var button = this.createButton(name, callback); + button.style.cssFloat = "left"; + button.style.marginRight = "4px"; + return button; + } + + createRightButton(name, callback) { + var button = this.createButton(name, callback); + button.style.cssFloat = "right"; + button.style.marginLeft = "4px"; + return button; + } + + createLeftSlider(self, name, callback) { + const divElement = document.createElement('div'); + divElement.id = "sam-confidence-slider"; + divElement.style.cssFloat = "left"; + divElement.style.fontFamily = "sans-serif"; + divElement.style.marginRight = "4px"; + divElement.style.color = "var(--input-text)"; + divElement.style.backgroundColor = "var(--comfy-input-bg)"; + divElement.style.borderRadius = "8px"; + divElement.style.borderColor = "var(--border-color)"; + divElement.style.borderStyle = "solid"; + divElement.style.fontSize = "15px"; + divElement.style.height = "21px"; + divElement.style.padding = "1px 6px"; + divElement.style.display = "flex"; + divElement.style.position = "relative"; + divElement.style.top = "2px"; + self.confidence_slider_input = document.createElement('input'); + self.confidence_slider_input.setAttribute('type', 'range'); + self.confidence_slider_input.setAttribute('min', '0'); + self.confidence_slider_input.setAttribute('max', '100'); + self.confidence_slider_input.setAttribute('value', '70'); + const labelElement = document.createElement("label"); + labelElement.textContent = name; + + divElement.appendChild(labelElement); + divElement.appendChild(self.confidence_slider_input); + + self.confidence_slider_input.addEventListener("change", callback); + + return divElement; + } + + async detect_and_invalidate_mask_canvas(self) { + const mask_img = await self.detect(self); + + const canvas = self.maskCtx.canvas; + const ctx = self.maskCtx; + + ctx.clearRect(0, 0, canvas.width, canvas.height); + + await new Promise((resolve, reject) => { + self.mask_image = new Image(); + self.mask_image.onload = function() { + ctx.drawImage(self.mask_image, 0, 0, canvas.width, canvas.height); + resolve(); + }; + self.mask_image.onerror = reject; + self.mask_image.src = mask_img.src; + }); + } + + setlayout(imgCanvas, maskCanvas, pointsCanvas) { + const self = this; + + // If it is specified as relative, using it only as a hidden placeholder for padding is recommended + // to prevent anomalies where it exceeds a certain size and goes outside of the window. + var placeholder = document.createElement("div"); + placeholder.style.position = "relative"; + placeholder.style.height = "50px"; + + var bottom_panel = document.createElement("div"); + bottom_panel.style.position = "absolute"; + bottom_panel.style.bottom = "0px"; + bottom_panel.style.left = "20px"; + bottom_panel.style.right = "20px"; + bottom_panel.style.height = "50px"; + + var brush = document.createElement("div"); + brush.id = "sam-brush"; + brush.style.backgroundColor = "blue"; + brush.style.outline = "2px solid pink"; + brush.style.borderRadius = "50%"; + brush.style.MozBorderRadius = "50%"; + brush.style.WebkitBorderRadius = "50%"; + brush.style.position = "absolute"; + brush.style.zIndex = 100; + brush.style.pointerEvents = "none"; + this.brush = brush; + this.element.appendChild(imgCanvas); + this.element.appendChild(maskCanvas); + this.element.appendChild(pointsCanvas); + this.element.appendChild(placeholder); // must below z-index than bottom_panel to avoid covering button + this.element.appendChild(bottom_panel); + document.body.appendChild(brush); + this.brush_size = 5; + + var confidence_slider = this.createLeftSlider(self, "Confidence", (event) => { + self.confidence = event.target.value; + }); + + var clearButton = this.createLeftButton("Clear", () => { + self.maskCtx.clearRect(0, 0, self.maskCanvas.width, self.maskCanvas.height); + self.pointsCtx.clearRect(0, 0, self.pointsCanvas.width, self.pointsCanvas.height); + + self.prompt_points = []; + + self.invalidatePointsCanvas(self); + }); + + var detectButton = this.createLeftButton("Detect", () => self.detect_and_invalidate_mask_canvas(self)); + + var cancelButton = this.createRightButton("Cancel", () => { + document.removeEventListener("mouseup", ImpactSamEditorDialog.handleMouseUp); + document.removeEventListener("keydown", ImpactSamEditorDialog.handleKeyDown); + self.close(); + }); + + self.saveButton = this.createRightButton("Save", () => { + document.removeEventListener("mouseup", ImpactSamEditorDialog.handleMouseUp); + document.removeEventListener("keydown", ImpactSamEditorDialog.handleKeyDown); + self.save(self); + }); + + var undoButton = this.createLeftButton("Undo", () => { + if(self.prompt_points.length > 0) { + self.prompt_points.pop(); + self.pointsCtx.clearRect(0, 0, self.pointsCanvas.width, self.pointsCanvas.height); + self.invalidatePointsCanvas(self); + } + }); + + bottom_panel.appendChild(clearButton); + bottom_panel.appendChild(detectButton); + bottom_panel.appendChild(self.saveButton); + bottom_panel.appendChild(cancelButton); + bottom_panel.appendChild(confidence_slider); + bottom_panel.appendChild(undoButton); + + imgCanvas.style.position = "relative"; + imgCanvas.style.top = "200"; + imgCanvas.style.left = "0"; + + maskCanvas.style.position = "absolute"; + maskCanvas.style.opacity = 0.5; + pointsCanvas.style.position = "absolute"; + } + + show() { + this.mask_image = null; + self.prompt_points = []; + + this.message_box = $el("p", ["Please wait a moment while the SAM model and the image are being loaded."]); + this.element.appendChild(this.message_box); + + if(self.imgCtx) { + self.imgCtx.clearRect(0, 0, self.imageCanvas.width, self.imageCanvas.height); + } + + const target_image_path = ComfyApp.clipspace.imgs[ComfyApp.clipspace['selectedIndex']].src; + this.load_sam(target_image_path); + + if(!this.is_layout_created) { + // layout + const imgCanvas = document.createElement('canvas'); + const maskCanvas = document.createElement('canvas'); + const pointsCanvas = document.createElement('canvas'); + + imgCanvas.id = "imageCanvas"; + maskCanvas.id = "maskCanvas"; + pointsCanvas.id = "pointsCanvas"; + + this.setlayout(imgCanvas, maskCanvas, pointsCanvas); + + // prepare content + this.imgCanvas = imgCanvas; + this.maskCanvas = maskCanvas; + this.pointsCanvas = pointsCanvas; + this.maskCtx = maskCanvas.getContext('2d'); + this.pointsCtx = pointsCanvas.getContext('2d'); + + this.is_layout_created = true; + + // replacement of onClose hook since close is not real close + const self = this; + const observer = new MutationObserver(function(mutations) { + mutations.forEach(function(mutation) { + if (mutation.type === 'attributes' && mutation.attributeName === 'style') { + if(self.last_display_style && self.last_display_style != 'none' && self.element.style.display == 'none') { + ComfyApp.onClipspaceEditorClosed(); + } + + self.last_display_style = self.element.style.display; + } + }); + }); + + const config = { attributes: true }; + observer.observe(this.element, config); + } + + this.setImages(target_image_path, this.imgCanvas, this.pointsCanvas); + + if(ComfyApp.clipspace_return_node) { + this.saveButton.innerText = "Save to node"; + } + else { + this.saveButton.innerText = "Save"; + } + this.saveButton.disabled = true; + + this.element.style.display = "block"; + this.element.style.zIndex = 8888; // NOTE: alert dialog must be high priority. + } + + updateBrushPreview(self, event) { + event.preventDefault(); + + const centerX = event.pageX; + const centerY = event.pageY; + + const brush = self.brush; + + brush.style.width = self.brush_size * 2 + "px"; + brush.style.height = self.brush_size * 2 + "px"; + brush.style.left = (centerX - self.brush_size) + "px"; + brush.style.top = (centerY - self.brush_size) + "px"; + } + + setImages(target_image_path, imgCanvas, pointsCanvas) { + const imgCtx = imgCanvas.getContext('2d'); + const maskCtx = this.maskCtx; + const maskCanvas = this.maskCanvas; + + const self = this; + + // image load + const orig_image = new Image(); + window.addEventListener("resize", () => { + // repositioning + imgCanvas.width = window.innerWidth - 250; + imgCanvas.height = window.innerHeight - 200; + + // redraw image + let drawWidth = orig_image.width; + let drawHeight = orig_image.height; + + if (orig_image.width > imgCanvas.width) { + drawWidth = imgCanvas.width; + drawHeight = (drawWidth / orig_image.width) * orig_image.height; + } + + if (drawHeight > imgCanvas.height) { + drawHeight = imgCanvas.height; + drawWidth = (drawHeight / orig_image.height) * orig_image.width; + } + + imgCtx.drawImage(orig_image, 0, 0, drawWidth, drawHeight); + + // update mask + pointsCanvas.width = drawWidth; + pointsCanvas.height = drawHeight; + pointsCanvas.style.top = imgCanvas.offsetTop + "px"; + pointsCanvas.style.left = imgCanvas.offsetLeft + "px"; + + maskCanvas.width = drawWidth; + maskCanvas.height = drawHeight; + maskCanvas.style.top = imgCanvas.offsetTop + "px"; + maskCanvas.style.left = imgCanvas.offsetLeft + "px"; + + self.invalidateMaskCanvas(self); + self.invalidatePointsCanvas(self); + }); + + // original image load + orig_image.onload = () => self.onLoaded(self); + const rgb_url = new URL(target_image_path); + rgb_url.searchParams.delete('channel'); + rgb_url.searchParams.set('channel', 'rgb'); + orig_image.src = rgb_url; + self.image = orig_image; + } + + onLoaded(self) { + if(self.message_box) { + self.element.removeChild(self.message_box); + self.message_box = null; + } + + window.dispatchEvent(new Event('resize')); + + self.setEventHandler(pointsCanvas); + self.saveButton.disabled = false; + } + + setEventHandler(targetCanvas) { + targetCanvas.addEventListener("contextmenu", (event) => { + event.preventDefault(); + }); + + const self = this; + targetCanvas.addEventListener('pointermove', (event) => this.updateBrushPreview(self,event)); + targetCanvas.addEventListener('pointerdown', (event) => this.handlePointerDown(self,event)); + targetCanvas.addEventListener('pointerover', (event) => { this.brush.style.display = "block"; }); + targetCanvas.addEventListener('pointerleave', (event) => { this.brush.style.display = "none"; }); + document.addEventListener('keydown', ImpactSamEditorDialog.handleKeyDown); + } + + static handleKeyDown(event) { + const self = ImpactSamEditorDialog.instance; + if (event.key === '=') { // positive + brush.style.backgroundColor = "blue"; + brush.style.outline = "2px solid pink"; + self.is_positive_mode = true; + } else if (event.key === '-') { // negative + brush.style.backgroundColor = "red"; + brush.style.outline = "2px solid skyblue"; + self.is_positive_mode = false; + } + } + + is_positive_mode = true; + prompt_points = []; + confidence = 70; + + invalidatePointsCanvas(self) { + const ctx = self.pointsCtx; + + for (const i in self.prompt_points) { + const [is_positive, x, y] = self.prompt_points[i]; + + const scaledX = x * ctx.canvas.width / self.image.width; + const scaledY = y * ctx.canvas.height / self.image.height; + + if(is_positive) + ctx.fillStyle = "blue"; + else + ctx.fillStyle = "red"; + ctx.beginPath(); + ctx.arc(scaledX, scaledY, 3, 0, 3 * Math.PI); + ctx.fill(); + } + } + + invalidateMaskCanvas(self) { + if(self.mask_image) { + self.maskCtx.clearRect(0, 0, self.maskCanvas.width, self.maskCanvas.height); + self.maskCtx.drawImage(self.mask_image, 0, 0, self.maskCanvas.width, self.maskCanvas.height); + } + } + + async load_sam(url) { + const parsedUrl = new URL(url); + const searchParams = new URLSearchParams(parsedUrl.search); + + const filename = searchParams.get("filename") || ""; + const fileType = searchParams.get("type") || ""; + const subfolder = searchParams.get("subfolder") || ""; + + const data = { + sam_model_name: "auto", + filename: filename, + type: fileType, + subfolder: subfolder + }; + + api.fetchApi('/sam/prepare', { + method: 'POST', + headers: { 'Content-Type': 'application/json' }, + body: JSON.stringify(data) + }); + } + + async detect(self) { + const positive_points = []; + const negative_points = []; + + for(const i in self.prompt_points) { + const [is_positive, x, y] = self.prompt_points[i]; + const point = [x,y]; + if(is_positive) + positive_points.push(point); + else + negative_points.push(point); + } + + const data = { + positive_points: positive_points, + negative_points: negative_points, + threshold: self.confidence/100 + }; + + const response = await api.fetchApi('/sam/detect', { + method: 'POST', + headers: { 'Content-Type': 'image/png' }, + body: JSON.stringify(data) + }); + + const blob = await response.blob(); + const url = URL.createObjectURL(blob); + + return new Promise((resolve, reject) => { + const image = new Image(); + image.onload = () => resolve(image); + image.onerror = reject; + image.src = url; + }); + } + + handlePointerDown(self, event) { + if ([0, 2, 5].includes(event.button)) { + event.preventDefault(); + const x = event.offsetX || event.targetTouches[0].clientX - maskRect.left; + const y = event.offsetY || event.targetTouches[0].clientY - maskRect.top; + + const originalX = x * self.image.width / self.pointsCanvas.width; + const originalY = y * self.image.height / self.pointsCanvas.height; + + var point = null; + if (event.button == 0) { + // positive + point = [true, originalX, originalY]; + } else { + // negative + point = [false, originalX, originalY]; + } + + self.prompt_points.push(point); + + self.invalidatePointsCanvas(self); + } + } + + async save(self) { + if(!self.mask_image) { + this.close(); + return; + } + + const save_canvas = document.createElement('canvas'); + + const save_ctx = save_canvas.getContext('2d', {willReadFrequently:true}); + save_canvas.width = self.mask_image.width; + save_canvas.height = self.mask_image.height; + + save_ctx.drawImage(self.mask_image, 0, 0, save_canvas.width, save_canvas.height); + + const save_data = save_ctx.getImageData(0, 0, save_canvas.width, save_canvas.height); + + // refine mask image + for (let i = 0; i < save_data.data.length; i += 4) { + if(save_data.data[i]) { + save_data.data[i+3] = 0; + } + else { + save_data.data[i+3] = 255; + } + + save_data.data[i] = 0; + save_data.data[i+1] = 0; + save_data.data[i+2] = 0; + } + + save_ctx.globalCompositeOperation = 'source-over'; + save_ctx.putImageData(save_data, 0, 0); + + const formData = new FormData(); + const filename = "clipspace-mask-" + performance.now() + ".png"; + + const item = + { + "filename": filename, + "subfolder": "", + "type": "temp", + }; + + if(ComfyApp.clipspace.images) + ComfyApp.clipspace.images[0] = item; + + if(ComfyApp.clipspace.widgets) { + const index = ComfyApp.clipspace.widgets.findIndex(obj => obj.name === 'image'); + + if(index >= 0) + ComfyApp.clipspace.widgets[index].value = `${filename} [temp]`; + } + + const dataURL = save_canvas.toDataURL(); + const blob = dataURLToBlob(dataURL); + + let original_url = new URL(this.image.src); + + const original_ref = { filename: original_url.searchParams.get('filename') }; + + let original_subfolder = original_url.searchParams.get("subfolder"); + if(original_subfolder) + original_ref.subfolder = original_subfolder; + + let original_type = original_url.searchParams.get("type"); + if(original_type) + original_ref.type = original_type; + + formData.append('image', blob, filename); + formData.append('original_ref', JSON.stringify(original_ref)); + formData.append('type', "temp"); + + await uploadMask(item, formData); + ComfyApp.onClipspaceEditorSave(); + this.close(); + } +} + +app.registerExtension({ + name: "Comfy.Impact.SAMEditor", + init(app) { + const callback = + function () { + let dlg = ImpactSamEditorDialog.getInstance(); + dlg.show(); + }; + + const context_predicate = () => ComfyApp.clipspace && ComfyApp.clipspace.imgs && ComfyApp.clipspace.imgs.length > 0 + ClipspaceDialog.registerButton("Impact SAM Detector", context_predicate, callback); + }, + + async beforeRegisterNodeDef(nodeType, nodeData, app) { + if (Array.isArray(nodeData.output) && (nodeData.output.includes("MASK") || nodeData.output.includes("IMAGE"))) { + addMenuHandler(nodeType, function (_, options) { + options.unshift({ + content: "Open in SAM Detector", + callback: () => { + ComfyApp.copyToClipspace(this); + ComfyApp.clipspace_return_node = this; + + let dlg = ImpactSamEditorDialog.getInstance(); + dlg.show(); + }, + }); + }); + } + } +}); + diff --git a/ComfyUI-Impact-Pack/js/impact-segs-picker.js b/ComfyUI-Impact-Pack/js/impact-segs-picker.js new file mode 100644 index 0000000000000000000000000000000000000000..16af55aa5af43e8ff523c3cfcdd8363f092e2d06 --- /dev/null +++ b/ComfyUI-Impact-Pack/js/impact-segs-picker.js @@ -0,0 +1,182 @@ +import { ComfyApp, app } from "../../scripts/app.js"; +import { ComfyDialog, $el } from "../../scripts/ui.js"; +import { api } from "../../scripts/api.js"; + +async function open_picker(node) { + const resp = await api.fetchApi(`/impact/segs/picker/count?id=${node.id}`); + const body = await resp.text(); + + let cnt = parseInt(body); + + var existingPicker = document.getElementById('impact-picker'); + if (existingPicker) { + existingPicker.parentNode.removeChild(existingPicker); + } + + var gallery = document.createElement('div'); + gallery.id = 'impact-picker'; + + gallery.style.position = "absolute"; + gallery.style.height = "80%"; + gallery.style.width = "80%"; + gallery.style.top = "10%"; + gallery.style.left = "10%"; + gallery.style.display = 'flex'; + gallery.style.flexWrap = 'wrap'; + gallery.style.maxHeight = '600px'; + gallery.style.overflow = 'auto'; + gallery.style.backgroundColor = 'rgba(0,0,0,0.3)'; + gallery.style.padding = '20px'; + gallery.draggable = false; + gallery.style.zIndex = 5000; + + var doneButton = document.createElement('button'); + doneButton.textContent = 'Done'; + doneButton.style.padding = '10px 10px'; + doneButton.style.border = 'none'; + doneButton.style.borderRadius = '5px'; + doneButton.style.fontFamily = 'Arial, sans-serif'; + doneButton.style.fontSize = '16px'; + doneButton.style.fontWeight = 'bold'; + doneButton.style.color = '#fff'; + doneButton.style.background = 'linear-gradient(to bottom, #0070B8, #003D66)'; + doneButton.style.boxShadow = '0 2px 4px rgba(0, 0, 0, 0.4)'; + doneButton.style.margin = "20px"; + doneButton.style.height = "40px"; + + var cancelButton = document.createElement('button'); + cancelButton.textContent = 'Cancel'; + cancelButton.style.padding = '10px 10px'; + cancelButton.style.border = 'none'; + cancelButton.style.borderRadius = '5px'; + cancelButton.style.fontFamily = 'Arial, sans-serif'; + cancelButton.style.fontSize = '16px'; + cancelButton.style.fontWeight = 'bold'; + cancelButton.style.color = '#fff'; + cancelButton.style.background = 'linear-gradient(to bottom, #ff70B8, #ff3D66)'; + cancelButton.style.boxShadow = '0 2px 4px rgba(0, 0, 0, 0.4)'; + cancelButton.style.margin = "20px"; + cancelButton.style.height = "40px"; + + const w = node.widgets.find((w) => w.name == 'picks'); + let prev_selected = w.value.split(',').map(function(item) { + return parseInt(item, 10); + }); + + let images = []; + doneButton.onclick = () => { + var result = ''; + for(let i in images) { + if(images[i].isSelected) { + if(result != '') + result += ', '; + + result += (parseInt(i)+1); + } + } + + w.value = result; + + gallery.parentNode.removeChild(gallery); + } + + cancelButton.onclick = () => { + gallery.parentNode.removeChild(gallery); + } + + var panel = document.createElement('div'); + panel.style.clear = 'both'; + panel.style.width = '100%'; + panel.style.height = '40px'; + panel.style.justifyContent = 'center'; + panel.style.alignItems = 'center'; + panel.style.display = 'flex'; + panel.appendChild(doneButton); + panel.appendChild(cancelButton); + gallery.appendChild(panel); + + var hint = document.createElement('label'); + hint.style.position = 'absolute'; + hint.innerHTML = 'Click: Toggle Selection
Ctrl-click: Single Selection'; + gallery.appendChild(hint); + + let max_size = 300; + + for(let i=0; i image.naturalHeight) { + ratio = max_size/image.naturalWidth; + } + else { + ratio = max_size/image.naturalHeight; + } + + let width = image.naturalWidth * ratio; + let height = image.naturalHeight * ratio; + + if(width < height) { + this.style.marginLeft = (200-width)/2+"px"; + } + else{ + this.style.marginTop = (200-height)/2+"px"; + } + + this.style.width = width+"px"; + this.style.height = height+"px"; + this.style.objectFit = 'cover'; + } + + image.addEventListener('click', function(event) { + if(event.ctrlKey) { + for(let i in images) { + if(images[i].isSelected) { + images[i].style.border = 'none'; + images[i].isSelected = false; + } + } + + image.style.border = '2px solid #006699'; + image.isSelected = true; + + return; + } + + if(image.isSelected) { + image.style.border = 'none'; + image.isSelected = false; + } + else { + image.style.border = '2px solid #006699'; + image.isSelected = true; + } + }); + + gallery.appendChild(image); + } + + document.body.appendChild(gallery); +} + + +app.registerExtension({ + name: "Comfy.Impack.Picker", + + nodeCreated(node, app) { + if(node.comfyClass == "ImpactSEGSPicker") { + node.addWidget("button", "pick", "image", () => { + open_picker(node); + }); + } + } +}); \ No newline at end of file diff --git a/ComfyUI-Impact-Pack/js/impact-wildcard.js b/ComfyUI-Impact-Pack/js/impact-wildcard.js new file mode 100644 index 0000000000000000000000000000000000000000..a3565135cd4758af84ba62aeb2ef23b56d8f606b --- /dev/null +++ b/ComfyUI-Impact-Pack/js/impact-wildcard.js @@ -0,0 +1,20 @@ +import { ComfyApp, app } from "../../scripts/app.js"; +import { api } from "../../scripts/api.js"; + +let refresh_btn = document.getElementById('comfy-refresh-button'); +let refresh_btn2 = document.querySelector('button[title="Refresh widgets in nodes to find new models or files"]'); + +let orig = refresh_btn.onclick; + +if(refresh_btn) { + refresh_btn.onclick = function() { + orig(); + api.fetchApi('/impact/wildcards/refresh'); + }; +} + +if(refresh_btn2) { + refresh_btn2?.addEventListener('click', function() { + api.fetchApi('/impact/wildcards/refresh'); + }); +} \ No newline at end of file diff --git a/ComfyUI-Impact-Pack/latent.png b/ComfyUI-Impact-Pack/latent.png new file mode 100644 index 0000000000000000000000000000000000000000..19fed324a25a7e1a2252400e7752ce5586742429 Binary files /dev/null and b/ComfyUI-Impact-Pack/latent.png differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/animatediff_nodes.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/animatediff_nodes.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..b5d3aeffd36351bf5931705bd8e901e50e64f054 Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/animatediff_nodes.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/bridge_nodes.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/bridge_nodes.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..388023d13dfcbc0465691ec828c4083741cb9a8b Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/bridge_nodes.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/config.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/config.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..727b33ac4732c10d81d749a7bd89d6e42d7d5aa3 Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/config.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/core.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/core.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..4cae40b70425c7b376c1bb71102e6ace23c28cf5 Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/core.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/defs.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/defs.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..e5f31cc01dae1df0cf2ca4867b312b7921894b43 Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/defs.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/detectors.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/detectors.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..b5d777050c210645dbed6dbef5336e3db3243a01 Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/detectors.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/hf_nodes.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/hf_nodes.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..b9b494e6c1ac3b6edcecd529a7bfc196ec9197cd Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/hf_nodes.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/hook_nodes.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/hook_nodes.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..0d67b9d65af8d2850bab7107e3a517dabc062856 Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/hook_nodes.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/hooks.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/hooks.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..512ce76df2c5925bfe18c00a7464f6a23d74c064 Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/hooks.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/impact_pack.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/impact_pack.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..a7faaf07183797eccf369a5edfeb41ef4e8ab02b Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/impact_pack.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/impact_sampling.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/impact_sampling.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..adcb3da079f7cf84fec0bd2363b93bac6782a510 Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/impact_sampling.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/impact_server.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/impact_server.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..7a7d34db3c538b69712538a33571585240f7b14c Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/impact_server.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/logics.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/logics.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..c2565004ddf7c952bd1a18613a96af362821512c Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/logics.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/pipe.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/pipe.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..970ab9889923ce4bbafaa5593953d6f83f24f57b Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/pipe.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/sample_error_enhancer.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/sample_error_enhancer.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..002d702a12b31c17d8d6b2ae79970f84e154f6d4 Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/sample_error_enhancer.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/segs_nodes.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/segs_nodes.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..c523c49d005fa27d28b1de7a6148ad224f419e62 Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/segs_nodes.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/segs_upscaler.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/segs_upscaler.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..252225b187b90fc63ec5a362923d28a2d1fb31e8 Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/segs_upscaler.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/special_samplers.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/special_samplers.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..9bcfae8ab0e80f840d9007c3bd1c744c84d17dbb Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/special_samplers.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/util_nodes.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/util_nodes.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..9f0a6e31f41000a883d81a886fd0576f806eb3a6 Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/util_nodes.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/utils.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/utils.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..3d24dd92842fbfb1d9b79abe114841a26565272a Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/utils.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/__pycache__/wildcards.cpython-312.pyc b/ComfyUI-Impact-Pack/modules/impact/__pycache__/wildcards.cpython-312.pyc new file mode 100644 index 0000000000000000000000000000000000000000..2827d24772b043977402589db2441dff578b5524 Binary files /dev/null and b/ComfyUI-Impact-Pack/modules/impact/__pycache__/wildcards.cpython-312.pyc differ diff --git a/ComfyUI-Impact-Pack/modules/impact/additional_dependencies.py b/ComfyUI-Impact-Pack/modules/impact/additional_dependencies.py new file mode 100644 index 0000000000000000000000000000000000000000..49d5f5c957502b66f7813127000d67b432cd14f2 --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/additional_dependencies.py @@ -0,0 +1,12 @@ +import sys +import subprocess + + +def ensure_onnx_package(): + try: + import onnxruntime + except Exception: + if "python_embeded" in sys.executable or "python_embedded" in sys.executable: + subprocess.check_call([sys.executable, '-s', '-m', 'pip', 'install', 'onnxruntime']) + else: + subprocess.check_call([sys.executable, '-s', '-m', 'pip', 'install', 'onnxruntime']) diff --git a/ComfyUI-Impact-Pack/modules/impact/animatediff_nodes.py b/ComfyUI-Impact-Pack/modules/impact/animatediff_nodes.py new file mode 100644 index 0000000000000000000000000000000000000000..acc62043624972620e98676b89804e197728d4f3 --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/animatediff_nodes.py @@ -0,0 +1,188 @@ +from nodes import MAX_RESOLUTION +from impact.utils import * +import impact.core as core +from impact.core import SEG +from impact.segs_nodes import SEGSPaste + + +try: + from comfy_extras import nodes_differential_diffusion +except Exception: + print(f"\n#############################################\n[Impact Pack] ComfyUI is an outdated version.\n#############################################\n") + raise Exception("[Impact Pack] ComfyUI is an outdated version.") + + +class SEGSDetailerForAnimateDiff: + @classmethod + def INPUT_TYPES(cls): + return {"required": { + "image_frames": ("IMAGE", ), + "segs": ("SEGS", ), + "guide_size": ("FLOAT", {"default": 512, "min": 64, "max": MAX_RESOLUTION, "step": 8}), + "guide_size_for": ("BOOLEAN", {"default": True, "label_on": "bbox", "label_off": "crop_region"}), + "max_size": ("FLOAT", {"default": 768, "min": 64, "max": MAX_RESOLUTION, "step": 8}), + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000}), + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS,), + "scheduler": (core.SCHEDULERS,), + "denoise": ("FLOAT", {"default": 0.5, "min": 0.0001, "max": 1.0, "step": 0.01}), + "basic_pipe": ("BASIC_PIPE",), + "refiner_ratio": ("FLOAT", {"default": 0.2, "min": 0.0, "max": 1.0}), + }, + "optional": { + "refiner_basic_pipe_opt": ("BASIC_PIPE",), + "noise_mask_feather": ("INT", {"default": 20, "min": 0, "max": 100, "step": 1}), + "scheduler_func_opt": ("SCHEDULER_FUNC",), + } + } + + RETURN_TYPES = ("SEGS", "IMAGE") + RETURN_NAMES = ("segs", "cnet_images") + OUTPUT_IS_LIST = (False, True) + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detailer" + + @staticmethod + def do_detail(image_frames, segs, guide_size, guide_size_for, max_size, seed, steps, cfg, sampler_name, scheduler, + denoise, basic_pipe, refiner_ratio=None, refiner_basic_pipe_opt=None, noise_mask_feather=0, scheduler_func_opt=None): + + model, clip, vae, positive, negative = basic_pipe + if refiner_basic_pipe_opt is None: + refiner_model, refiner_clip, refiner_positive, refiner_negative = None, None, None, None + else: + refiner_model, refiner_clip, _, refiner_positive, refiner_negative = refiner_basic_pipe_opt + + segs = core.segs_scale_match(segs, image_frames.shape) + + new_segs = [] + cnet_image_list = [] + + if noise_mask_feather > 0 and 'denoise_mask_function' not in model.model_options: + model = nodes_differential_diffusion.DifferentialDiffusion().apply(model)[0] + + for seg in segs[1]: + cropped_image_frames = None + + for image in image_frames: + image = image.unsqueeze(0) + cropped_image = seg.cropped_image if seg.cropped_image is not None else crop_tensor4(image, seg.crop_region) + cropped_image = to_tensor(cropped_image) + if cropped_image_frames is None: + cropped_image_frames = cropped_image + else: + cropped_image_frames = torch.concat((cropped_image_frames, cropped_image), dim=0) + + cropped_image_frames = cropped_image_frames.cpu().numpy() + + # It is assumed that AnimateDiff does not support conditioning masks based on test results, but it will be added for future consideration. + cropped_positive = [ + [condition, { + k: core.crop_condition_mask(v, cropped_image_frames, seg.crop_region) if k == "mask" else v + for k, v in details.items() + }] + for condition, details in positive + ] + + cropped_negative = [ + [condition, { + k: core.crop_condition_mask(v, cropped_image_frames, seg.crop_region) if k == "mask" else v + for k, v in details.items() + }] + for condition, details in negative + ] + + enhanced_image_tensor, cnet_images = core.enhance_detail_for_animatediff(cropped_image_frames, model, clip, vae, guide_size, guide_size_for, max_size, + seg.bbox, seed, steps, cfg, sampler_name, scheduler, + cropped_positive, cropped_negative, denoise, seg.cropped_mask, + refiner_ratio=refiner_ratio, refiner_model=refiner_model, + refiner_clip=refiner_clip, refiner_positive=refiner_positive, + refiner_negative=refiner_negative, control_net_wrapper=seg.control_net_wrapper, + noise_mask_feather=noise_mask_feather, scheduler_func=scheduler_func_opt) + if cnet_images is not None: + cnet_image_list.extend(cnet_images) + + if enhanced_image_tensor is None: + new_cropped_image = cropped_image_frames + else: + new_cropped_image = enhanced_image_tensor.cpu().numpy() + + new_seg = SEG(new_cropped_image, seg.cropped_mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, None) + new_segs.append(new_seg) + + return (segs[0], new_segs), cnet_image_list + + def doit(self, image_frames, segs, guide_size, guide_size_for, max_size, seed, steps, cfg, sampler_name, scheduler, + denoise, basic_pipe, refiner_ratio=None, refiner_basic_pipe_opt=None, inpaint_model=False, noise_mask_feather=0, scheduler_func_opt=None): + + segs, cnet_images = SEGSDetailerForAnimateDiff.do_detail(image_frames, segs, guide_size, guide_size_for, max_size, seed, steps, cfg, sampler_name, + scheduler, denoise, basic_pipe, refiner_ratio, refiner_basic_pipe_opt, + noise_mask_feather=noise_mask_feather, scheduler_func_opt=scheduler_func_opt) + + if len(cnet_images) == 0: + cnet_images = [empty_pil_tensor()] + + return (segs, cnet_images) + + +class DetailerForEachPipeForAnimateDiff: + @classmethod + def INPUT_TYPES(cls): + return {"required": { + "image_frames": ("IMAGE", ), + "segs": ("SEGS", ), + "guide_size": ("FLOAT", {"default": 512, "min": 64, "max": nodes.MAX_RESOLUTION, "step": 8}), + "guide_size_for": ("BOOLEAN", {"default": True, "label_on": "bbox", "label_off": "crop_region"}), + "max_size": ("FLOAT", {"default": 1024, "min": 64, "max": nodes.MAX_RESOLUTION, "step": 8}), + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000}), + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS,), + "scheduler": (core.SCHEDULERS,), + "denoise": ("FLOAT", {"default": 0.5, "min": 0.0001, "max": 1.0, "step": 0.01}), + "feather": ("INT", {"default": 5, "min": 0, "max": 100, "step": 1}), + "basic_pipe": ("BASIC_PIPE", ), + "refiner_ratio": ("FLOAT", {"default": 0.2, "min": 0.0, "max": 1.0}), + }, + "optional": { + "detailer_hook": ("DETAILER_HOOK",), + "refiner_basic_pipe_opt": ("BASIC_PIPE",), + "noise_mask_feather": ("INT", {"default": 20, "min": 0, "max": 100, "step": 1}), + "scheduler_func_opt": ("SCHEDULER_FUNC",), + } + } + + RETURN_TYPES = ("IMAGE", "SEGS", "BASIC_PIPE", "IMAGE") + RETURN_NAMES = ("image", "segs", "basic_pipe", "cnet_images") + OUTPUT_IS_LIST = (False, False, False, True) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detailer" + + @staticmethod + def doit(image_frames, segs, guide_size, guide_size_for, max_size, seed, steps, cfg, sampler_name, scheduler, + denoise, feather, basic_pipe, refiner_ratio=None, detailer_hook=None, refiner_basic_pipe_opt=None, + noise_mask_feather=0, scheduler_func_opt=None): + + enhanced_segs = [] + cnet_image_list = [] + + for sub_seg in segs[1]: + single_seg = segs[0], [sub_seg] + enhanced_seg, cnet_images = SEGSDetailerForAnimateDiff().do_detail(image_frames, single_seg, guide_size, guide_size_for, max_size, seed, steps, cfg, sampler_name, scheduler, + denoise, basic_pipe, refiner_ratio, refiner_basic_pipe_opt, noise_mask_feather, scheduler_func_opt=scheduler_func_opt) + + image_frames = SEGSPaste.doit(image_frames, enhanced_seg, feather, alpha=255)[0] + + if cnet_images is not None: + cnet_image_list.extend(cnet_images) + + if detailer_hook is not None: + image_frames = detailer_hook.post_paste(image_frames) + + enhanced_segs += enhanced_seg[1] + + new_segs = segs[0], enhanced_segs + return image_frames, new_segs, basic_pipe, cnet_image_list diff --git a/ComfyUI-Impact-Pack/modules/impact/bridge_nodes.py b/ComfyUI-Impact-Pack/modules/impact/bridge_nodes.py new file mode 100644 index 0000000000000000000000000000000000000000..1d3af15a1e134dccafe7826bcfd2d082ab44c352 --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/bridge_nodes.py @@ -0,0 +1,380 @@ +import os +from PIL import ImageOps +from impact.utils import * +import latent_preview + + +# NOTE: this should not be `from . import core`. +# I don't know why but... 'from .' and 'from impact' refer to different core modules. +# This separates global variables of the core module and breaks the preview bridge. +from impact import core +# <-- +import random + + +class PreviewBridge: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "images": ("IMAGE",), + "image": ("STRING", {"default": ""}), + }, + "optional": { + "block": ("BOOLEAN", {"default": False, "label_on": "if_empty_mask", "label_off": "never", "tooltip": "is_empty_mask: If the mask is empty, the execution is stopped.\nnever: The execution is never stopped."}), + "restore_mask": (["never", "always", "if_same_size"], {"tooltip": "if_same_size: If the changed input image is the same size as the previous image, restore using the last saved mask\nalways: Whenever the input image changes, always restore using the last saved mask\nnever: Do not restore the mask.\n`restore_mask` has higher priority than `block`"}), + }, + "hidden": {"unique_id": "UNIQUE_ID", "extra_pnginfo": "EXTRA_PNGINFO"}, + } + + RETURN_TYPES = ("IMAGE", "MASK", ) + + FUNCTION = "doit" + + OUTPUT_NODE = True + + CATEGORY = "ImpactPack/Util" + + DESCRIPTION = "This is a feature that allows you to edit and send a Mask over a image.\nIf the block is set to 'is_empty_mask', the execution is stopped when the mask is empty." + + def __init__(self): + super().__init__() + self.output_dir = folder_paths.get_temp_directory() + self.type = "temp" + self.prev_hash = None + + @staticmethod + def load_image(pb_id): + is_fail = False + if pb_id not in core.preview_bridge_image_id_map: + is_fail = True + + image_path, ui_item = core.preview_bridge_image_id_map[pb_id] + + if not os.path.isfile(image_path): + is_fail = True + + if not is_fail: + i = Image.open(image_path) + i = ImageOps.exif_transpose(i) + image = i.convert("RGB") + image = np.array(image).astype(np.float32) / 255.0 + image = torch.from_numpy(image)[None,] + + if 'A' in i.getbands(): + mask = np.array(i.getchannel('A')).astype(np.float32) / 255.0 + mask = 1. - torch.from_numpy(mask) + else: + mask = torch.zeros((64, 64), dtype=torch.float32, device="cpu") + else: + image = empty_pil_tensor() + mask = torch.zeros((64, 64), dtype=torch.float32, device="cpu") + ui_item = { + "filename": 'empty.png', + "subfolder": '', + "type": 'temp' + } + + return image, mask.unsqueeze(0), ui_item + + def doit(self, images, image, unique_id, block=False, restore_mask="never", prompt=None, extra_pnginfo=None): + need_refresh = False + + if unique_id not in core.preview_bridge_cache: + need_refresh = True + + elif core.preview_bridge_cache[unique_id][0] is not images: + need_refresh = True + + if not need_refresh: + pixels, mask, path_item = PreviewBridge.load_image(image) + image = [path_item] + else: + if restore_mask != "never": + mask = core.preview_bridge_last_mask_cache.get(unique_id) + if mask is None or (restore_mask != "always" and mask.shape[1:] != images.shape[1:3]): + mask = None + else: + mask = None + + if mask is None: + mask = torch.zeros((64, 64), dtype=torch.float32, device="cpu") + res = nodes.PreviewImage().save_images(images, filename_prefix="PreviewBridge/PB-", prompt=prompt, extra_pnginfo=extra_pnginfo) + else: + masked_images = tensor_convert_rgba(images) + resized_mask = resize_mask(mask, (images.shape[1], images.shape[2])).unsqueeze(3) + resized_mask = 1 - resized_mask + tensor_putalpha(masked_images, resized_mask) + res = nodes.PreviewImage().save_images(masked_images, filename_prefix="PreviewBridge/PB-", prompt=prompt, extra_pnginfo=extra_pnginfo) + + image2 = res['ui']['images'] + pixels = images + + path = os.path.join(folder_paths.get_temp_directory(), 'PreviewBridge', image2[0]['filename']) + core.set_previewbridge_image(unique_id, path, image2[0]) + core.preview_bridge_image_id_map[image] = (path, image2[0]) + core.preview_bridge_image_name_map[unique_id, path] = (image, image2[0]) + core.preview_bridge_cache[unique_id] = (images, image2) + + image = image2 + + is_empty_mask = torch.all(mask == 0) + + if block and is_empty_mask and core.is_execution_model_version_supported(): + from comfy_execution.graph import ExecutionBlocker + result = ExecutionBlocker(None), ExecutionBlocker(None) + elif block and is_empty_mask: + print(f"[Impact Pack] PreviewBridge: ComfyUI is outdated - blocking feature is disabled.") + result = pixels, mask + else: + result = pixels, mask + + if not is_empty_mask: + core.preview_bridge_last_mask_cache[unique_id] = mask + + return { + "ui": {"images": image}, + "result": result, + } + + +def decode_latent(latent, preview_method, vae_opt=None): + if vae_opt is not None: + image = nodes.VAEDecode().decode(vae_opt, latent)[0] + return image + + from comfy.cli_args import LatentPreviewMethod + import comfy.latent_formats as latent_formats + + if preview_method.startswith("TAE"): + decoder_name = None + + if preview_method == "TAESD15": + decoder_name = "taesd" + elif preview_method == 'TAESDXL': + decoder_name = "taesdxl" + elif preview_method == 'TAESD3': + decoder_name = "taesd3" + elif preview_method == 'TAEF1': + decoder_name = "taef1" + + if decoder_name: + vae = nodes.VAELoader().load_vae(decoder_name)[0] + image = nodes.VAEDecode().decode(vae, latent)[0] + return image + + if preview_method == "Latent2RGB-SD15": + latent_format = latent_formats.SD15() + method = LatentPreviewMethod.Latent2RGB + elif preview_method == "Latent2RGB-SDXL": + latent_format = latent_formats.SDXL() + method = LatentPreviewMethod.Latent2RGB + elif preview_method == "Latent2RGB-SD3": + latent_format = latent_formats.SD3() + method = LatentPreviewMethod.Latent2RGB + elif preview_method == "Latent2RGB-SD-X4": + latent_format = latent_formats.SD_X4() + method = LatentPreviewMethod.Latent2RGB + elif preview_method == "Latent2RGB-Playground-2.5": + latent_format = latent_formats.SDXL_Playground_2_5() + method = LatentPreviewMethod.Latent2RGB + elif preview_method == "Latent2RGB-SC-Prior": + latent_format = latent_formats.SC_Prior() + method = LatentPreviewMethod.Latent2RGB + elif preview_method == "Latent2RGB-SC-B": + latent_format = latent_formats.SC_B() + method = LatentPreviewMethod.Latent2RGB + elif preview_method == "Latent2RGB-FLUX.1": + latent_format = latent_formats.Flux() + method = LatentPreviewMethod.Latent2RGB + else: + print(f"[Impact Pack] PreviewBridgeLatent: '{preview_method}' is unsupported preview method.") + latent_format = latent_formats.SD15() + method = LatentPreviewMethod.Latent2RGB + + previewer = core.get_previewer("cpu", latent_format=latent_format, force=True, method=method) + samples = latent_format.process_in(latent['samples']) + + pil_image = previewer.decode_latent_to_preview(samples) + pixels_size = pil_image.size[0]*8, pil_image.size[1]*8 + resized_image = pil_image.resize(pixels_size, resample=LANCZOS) + + return to_tensor(resized_image).unsqueeze(0) + + +class PreviewBridgeLatent: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "latent": ("LATENT",), + "image": ("STRING", {"default": ""}), + "preview_method": (["Latent2RGB-FLUX.1", + "Latent2RGB-SDXL", "Latent2RGB-SD15", "Latent2RGB-SD3", + "Latent2RGB-SD-X4", "Latent2RGB-Playground-2.5", + "Latent2RGB-SC-Prior", "Latent2RGB-SC-B", + "TAEF1", "TAESDXL", "TAESD15", "TAESD3"],), + }, + "optional": { + "vae_opt": ("VAE", ), + "block": ("BOOLEAN", {"default": False, "label_on": "if_empty_mask", "label_off": "never", "tooltip": "is_empty_mask: If the mask is empty, the execution is stopped.\nnever: The execution is never stopped. Instead, it returns a white mask."}), + "restore_mask": (["never", "always", "if_same_size"], {"tooltip": "if_same_size: If the changed input latent is the same size as the previous latent, restore using the last saved mask\nalways: Whenever the input latent changes, always restore using the last saved mask\nnever: Do not restore the mask.\n`restore_mask` has higher priority than `block`\nIf the input latent already has a mask, do not restore mask."}), + }, + "hidden": {"unique_id": "UNIQUE_ID", "prompt": "PROMPT", "extra_pnginfo": "EXTRA_PNGINFO"}, + } + + RETURN_TYPES = ("LATENT", "MASK", ) + + FUNCTION = "doit" + + OUTPUT_NODE = True + + CATEGORY = "ImpactPack/Util" + + DESCRIPTION = "This is a feature that allows you to edit and send a Mask over a latent image.\nIf the block is set to 'is_empty_mask', the execution is stopped when the mask is empty." + + def __init__(self): + super().__init__() + self.output_dir = folder_paths.get_temp_directory() + self.type = "temp" + self.prev_hash = None + self.prefix_append = "_temp_" + ''.join(random.choice("abcdefghijklmnopqrstupvxyz") for x in range(5)) + + @staticmethod + def load_image(pb_id): + is_fail = False + if pb_id not in core.preview_bridge_image_id_map: + is_fail = True + + image_path, ui_item = core.preview_bridge_image_id_map[pb_id] + + if not os.path.isfile(image_path): + is_fail = True + + if not is_fail: + i = Image.open(image_path) + i = ImageOps.exif_transpose(i) + image = i.convert("RGB") + image = np.array(image).astype(np.float32) / 255.0 + image = torch.from_numpy(image)[None,] + + if 'A' in i.getbands(): + mask = np.array(i.getchannel('A')).astype(np.float32) / 255.0 + mask = 1. - torch.from_numpy(mask) + else: + mask = None + else: + image = empty_pil_tensor() + mask = None + ui_item = { + "filename": 'empty.png', + "subfolder": '', + "type": 'temp' + } + + return image, mask, ui_item + + def doit(self, latent, image, preview_method, vae_opt=None, block=False, unique_id=None, restore_mask='never', prompt=None, extra_pnginfo=None): + latent_channels = latent['samples'].shape[1] + preview_method_channels = 16 if 'SD3' in preview_method or 'SC-Prior' in preview_method or 'FLUX.1' in preview_method or 'TAEF1' == preview_method else 4 + + if vae_opt is None and latent_channels != preview_method_channels: + print(f"[PreviewBridgeLatent] The version of latent is not compatible with preview_method.\nSD3, SD1/SD2, SDXL, SC-Prior, SC-B and FLUX.1 are not compatible with each other.") + raise Exception("The version of latent is not compatible with preview_method.
SD3, SD1/SD2, SDXL, SC-Prior, SC-B and FLUX.1 are not compatible with each other.") + + need_refresh = False + + if unique_id not in core.preview_bridge_cache: + need_refresh = True + + elif (core.preview_bridge_cache[unique_id][0] is not latent + or (vae_opt is None and core.preview_bridge_cache[unique_id][2] is not None) + or (vae_opt is None and core.preview_bridge_cache[unique_id][1] != preview_method) + or (vae_opt is not None and core.preview_bridge_cache[unique_id][2] is not vae_opt)): + need_refresh = True + + if not need_refresh: + pixels, mask, path_item = PreviewBridge.load_image(image) + + if mask is None: + mask = torch.ones(latent['samples'].shape[2:], dtype=torch.float32, device="cpu").unsqueeze(0) + if 'noise_mask' in latent: + res_latent = latent.copy() + del res_latent['noise_mask'] + else: + res_latent = latent + + is_empty_mask = True + else: + res_latent = latent.copy() + res_latent['noise_mask'] = mask + + is_empty_mask = torch.all(mask == 1) + + res_image = [path_item] + else: + decoded_image = decode_latent(latent, preview_method, vae_opt) + + if 'noise_mask' in latent: + mask = latent['noise_mask'].squeeze(0) # 4D mask -> 3D mask + + decoded_pil = to_pil(decoded_image) + + inverted_mask = 1 - mask # invert + resized_mask = resize_mask(inverted_mask, (decoded_image.shape[1], decoded_image.shape[2])) + result_pil = apply_mask_alpha_to_pil(decoded_pil, resized_mask) + + full_output_folder, filename, counter, _, _ = folder_paths.get_save_image_path("PreviewBridge/PBL-"+self.prefix_append, folder_paths.get_temp_directory(), result_pil.size[0], result_pil.size[1]) + file = f"{filename}_{counter}.png" + result_pil.save(os.path.join(full_output_folder, file), compress_level=4) + res_image = [{ + 'filename': file, + 'subfolder': 'PreviewBridge', + 'type': 'temp', + }] + + is_empty_mask = False + else: + if restore_mask != "never": + mask = core.preview_bridge_last_mask_cache.get(unique_id) + if mask is None or (restore_mask != "always" and mask.shape[1:] != decoded_image.shape[1:3]): + mask = None + else: + mask = None + + if mask is None: + mask = torch.ones(latent['samples'].shape[2:], dtype=torch.float32, device="cpu").unsqueeze(0) + res = nodes.PreviewImage().save_images(decoded_image, filename_prefix="PreviewBridge/PBL-", prompt=prompt, extra_pnginfo=extra_pnginfo) + else: + masked_images = tensor_convert_rgba(decoded_image) + resized_mask = resize_mask(mask, (decoded_image.shape[1], decoded_image.shape[2])).unsqueeze(3) + resized_mask = 1 - resized_mask + tensor_putalpha(masked_images, resized_mask) + res = nodes.PreviewImage().save_images(masked_images, filename_prefix="PreviewBridge/PBL-", prompt=prompt, extra_pnginfo=extra_pnginfo) + + res_image = res['ui']['images'] + + is_empty_mask = torch.all(mask == 1) + + path = os.path.join(folder_paths.get_temp_directory(), 'PreviewBridge', res_image[0]['filename']) + core.set_previewbridge_image(unique_id, path, res_image[0]) + core.preview_bridge_image_id_map[image] = (path, res_image[0]) + core.preview_bridge_image_name_map[unique_id, path] = (image, res_image[0]) + core.preview_bridge_cache[unique_id] = (latent, preview_method, vae_opt, res_image) + + res_latent = latent + + if block and is_empty_mask and core.is_execution_model_version_supported(): + from comfy_execution.graph import ExecutionBlocker + result = ExecutionBlocker(None), ExecutionBlocker(None) + elif block and is_empty_mask: + print(f"[Impact Pack] PreviewBridgeLatent: ComfyUI is outdated - blocking feature is disabled.") + result = res_latent, mask + else: + result = res_latent, mask + + if not is_empty_mask: + core.preview_bridge_last_mask_cache[unique_id] = mask + + return { + "ui": {"images": res_image}, + "result": result, + } diff --git a/ComfyUI-Impact-Pack/modules/impact/config.py b/ComfyUI-Impact-Pack/modules/impact/config.py new file mode 100644 index 0000000000000000000000000000000000000000..b337f6f9cc738342e00e9f2d7a0d78a2a19d832c --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/config.py @@ -0,0 +1,68 @@ +import configparser +import os + +version_code = [7, 10, 7] +version = f"V{version_code[0]}.{version_code[1]}" + (f'.{version_code[2]}' if len(version_code) > 2 else '') + +dependency_version = 23 + +my_path = os.path.dirname(__file__) +old_config_path = os.path.join(my_path, "impact-pack.ini") +config_path = os.path.join(my_path, "..", "..", "impact-pack.ini") +latent_letter_path = os.path.join(my_path, "..", "..", "latent.png") + + +def write_config(): + config = configparser.ConfigParser() + config['default'] = { + 'dependency_version': str(dependency_version), + 'mmdet_skip': str(get_config()['mmdet_skip']), + 'sam_editor_cpu': str(get_config()['sam_editor_cpu']), + 'sam_editor_model': get_config()['sam_editor_model'], + 'custom_wildcards': get_config()['custom_wildcards'], + 'disable_gpu_opencv': get_config()['disable_gpu_opencv'], + } + with open(config_path, 'w') as configfile: + config.write(configfile) + + +def read_config(): + try: + config = configparser.ConfigParser() + config.read(config_path) + default_conf = config['default'] + + if not os.path.exists(default_conf['custom_wildcards']): + print(f"[WARN] ComfyUI-Impact-Pack: custom_wildcards path not found: {default_conf['custom_wildcards']}. Using default path.") + default_conf['custom_wildcards'] = os.path.join(my_path, "..", "..", "custom_wildcards") + + return { + 'dependency_version': int(default_conf['dependency_version']), + 'mmdet_skip': default_conf['mmdet_skip'].lower() == 'true' if 'mmdet_skip' in default_conf else True, + 'sam_editor_cpu': default_conf['sam_editor_cpu'].lower() == 'true' if 'sam_editor_cpu' in default_conf else False, + 'sam_editor_model': default_conf['sam_editor_model'].lower() if 'sam_editor_model' else 'sam_vit_b_01ec64.pth', + 'custom_wildcards': default_conf['custom_wildcards'] if 'custom_wildcards' in default_conf else os.path.abspath(os.path.join(os.path.dirname(__file__), "..", "..", "custom_wildcards")), + 'disable_gpu_opencv': default_conf['disable_gpu_opencv'].lower() == 'true' if 'disable_gpu_opencv' in default_conf else True + } + + except Exception: + return { + 'dependency_version': 0, + 'mmdet_skip': True, + 'sam_editor_cpu': False, + 'sam_editor_model': 'sam_vit_b_01ec64.pth', + 'custom_wildcards': os.path.abspath(os.path.join(os.path.dirname(__file__), "..", "..", "custom_wildcards")), + 'disable_gpu_opencv': True + } + + +cached_config = None + + +def get_config(): + global cached_config + + if cached_config is None: + cached_config = read_config() + + return cached_config diff --git a/ComfyUI-Impact-Pack/modules/impact/core.py b/ComfyUI-Impact-Pack/modules/impact/core.py new file mode 100644 index 0000000000000000000000000000000000000000..a31c3a2fe00b2e8f70a39f8cd726afadc8690ca7 --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/core.py @@ -0,0 +1,2208 @@ +import copy +import os +import warnings + +import numpy +import torch +from segment_anything import SamPredictor + +from comfy_extras.nodes_custom_sampler import Noise_RandomNoise +from impact.utils import * +from collections import namedtuple +import numpy as np +from skimage.measure import label + +import nodes +import comfy_extras.nodes_upscale_model as model_upscale +from server import PromptServer +import comfy +import impact.wildcards as wildcards +import math +import cv2 +import time +from comfy import model_management +from impact import utils +from impact import impact_sampling +from concurrent.futures import ThreadPoolExecutor +import inspect + + +try: + from comfy_extras import nodes_differential_diffusion +except Exception: + print(f"\n#############################################\n[Impact Pack] ComfyUI is an outdated version.\n#############################################\n") + raise Exception("[Impact Pack] ComfyUI is an outdated version.") + + +SEG = namedtuple("SEG", + ['cropped_image', 'cropped_mask', 'confidence', 'crop_region', 'bbox', 'label', 'control_net_wrapper'], + defaults=[None]) + +pb_id_cnt = time.time() +preview_bridge_image_id_map = {} +preview_bridge_image_name_map = {} + +preview_bridge_cache = {} +preview_bridge_last_mask_cache = {} + +current_prompt = None + +SCHEDULERS = comfy.samplers.KSampler.SCHEDULERS + ['AYS SDXL', 'AYS SD1', 'AYS SVD', 'GITS[coeff=1.2]'] + + +def is_execution_model_version_supported(): + try: + import comfy_execution + return True + except: + return False + + +def set_previewbridge_image(node_id, file, item): + global pb_id_cnt + + if file in preview_bridge_image_name_map: + pb_id = preview_bridge_image_name_map[node_id, file] + if pb_id.startswith(f"${node_id}"): + return pb_id + + pb_id = f"${node_id}-{pb_id_cnt}" + preview_bridge_image_id_map[pb_id] = (file, item) + preview_bridge_image_name_map[node_id, file] = (pb_id, item) + pb_id_cnt += 1 + + return pb_id + + +def erosion_mask(mask, grow_mask_by): + mask = make_2d_mask(mask) + + w = mask.shape[1] + h = mask.shape[0] + + device = comfy.model_management.get_torch_device() + mask = mask.clone().to(device) + mask2 = torch.nn.functional.interpolate(mask.reshape((-1, 1, mask.shape[-2], mask.shape[-1])), size=(w, h), mode="bilinear").to(device) + if grow_mask_by == 0: + mask_erosion = mask2 + else: + kernel_tensor = torch.ones((1, 1, grow_mask_by, grow_mask_by)).to(device) + padding = math.ceil((grow_mask_by - 1) / 2) + + mask_erosion = torch.clamp(torch.nn.functional.conv2d(mask2.round(), kernel_tensor, padding=padding), 0, 1) + + return mask_erosion[:, :, :w, :h].round().cpu() + + +# CREDIT: https://github.com/BlenderNeko/ComfyUI_Noise/blob/afb14757216257b12268c91845eac248727a55e2/nodes.py#L68 +# https://discuss.pytorch.org/t/help-regarding-slerp-function-for-generative-model-sampling/32475/3 +def slerp(val, low, high): + dims = low.shape + + low = low.reshape(dims[0], -1) + high = high.reshape(dims[0], -1) + + low_norm = low/torch.norm(low, dim=1, keepdim=True) + high_norm = high/torch.norm(high, dim=1, keepdim=True) + + low_norm[low_norm != low_norm] = 0.0 + high_norm[high_norm != high_norm] = 0.0 + + omega = torch.acos((low_norm*high_norm).sum(1)) + so = torch.sin(omega) + res = (torch.sin((1.0-val)*omega)/so).unsqueeze(1)*low + (torch.sin(val*omega)/so).unsqueeze(1) * high + + return res.reshape(dims) + + +def mix_noise(from_noise, to_noise, strength, variation_method): + if variation_method == 'slerp': + mixed_noise = slerp(strength, from_noise, to_noise) + else: + # linear + mixed_noise = (1 - strength) * from_noise + strength * to_noise + + # NOTE: Since the variance of the Gaussian noise in mixed_noise has changed, it must be corrected through scaling. + scale_factor = math.sqrt((1 - strength) ** 2 + strength ** 2) + mixed_noise /= scale_factor + + return mixed_noise + + +class REGIONAL_PROMPT: + def __init__(self, mask, sampler, variation_seed=0, variation_strength=0.0, variation_method='linear'): + mask = make_2d_mask(mask) + + self.mask = mask + self.sampler = sampler + self.mask_erosion = None + self.erosion_factor = None + self.variation_seed = variation_seed + self.variation_strength = variation_strength + self.variation_method = variation_method + + def clone_with_sampler(self, sampler): + rp = REGIONAL_PROMPT(self.mask, sampler) + rp.mask_erosion = self.mask_erosion + rp.erosion_factor = self.erosion_factor + rp.variation_seed = self.variation_seed + rp.variation_strength = self.variation_strength + rp.variation_method = self.variation_method + return rp + + def get_mask_erosion(self, factor): + if self.mask_erosion is None or self.erosion_factor != factor: + self.mask_erosion = erosion_mask(self.mask, factor) + self.erosion_factor = factor + + return self.mask_erosion + + def touch_noise(self, noise): + if self.variation_strength > 0.0: + mask = utils.make_3d_mask(self.mask) + mask = utils.resize_mask(mask, (noise.shape[2], noise.shape[3])).unsqueeze(0) + + regional_noise = Noise_RandomNoise(self.variation_seed).generate_noise({'samples': noise}) + mixed_noise = mix_noise(noise, regional_noise, self.variation_strength, variation_method=self.variation_method) + + return (mask == 1).float() * mixed_noise + (mask == 0).float() * noise + + return noise + + +class NO_BBOX_DETECTOR: + pass + + +class NO_SEGM_DETECTOR: + pass + + +def create_segmasks(results): + bboxs = results[1] + segms = results[2] + confidence = results[3] + + results = [] + for i in range(len(segms)): + item = (bboxs[i], segms[i].astype(np.float32), confidence[i]) + results.append(item) + return results + + +def gen_detection_hints_from_mask_area(x, y, mask, threshold, use_negative): + mask = make_2d_mask(mask) + + points = [] + plabs = [] + + # minimum sampling step >= 3 + y_step = max(3, int(mask.shape[0] / 20)) + x_step = max(3, int(mask.shape[1] / 20)) + + for i in range(0, len(mask), y_step): + for j in range(0, len(mask[i]), x_step): + if mask[i][j] > threshold: + points.append((x + j, y + i)) + plabs.append(1) + elif use_negative and mask[i][j] == 0: + points.append((x + j, y + i)) + plabs.append(0) + + return points, plabs + + +def gen_negative_hints(w, h, x1, y1, x2, y2): + npoints = [] + nplabs = [] + + # minimum sampling step >= 3 + y_step = max(3, int(w / 20)) + x_step = max(3, int(h / 20)) + + for i in range(10, h - 10, y_step): + for j in range(10, w - 10, x_step): + if not (x1 - 10 <= j and j <= x2 + 10 and y1 - 10 <= i and i <= y2 + 10): + npoints.append((j, i)) + nplabs.append(0) + + return npoints, nplabs + + +def enhance_detail(image, model, clip, vae, guide_size, guide_size_for_bbox, max_size, bbox, seed, steps, cfg, + sampler_name, + scheduler, positive, negative, denoise, noise_mask, force_inpaint, + wildcard_opt=None, wildcard_opt_concat_mode=None, + detailer_hook=None, + refiner_ratio=None, refiner_model=None, refiner_clip=None, refiner_positive=None, + refiner_negative=None, control_net_wrapper=None, cycle=1, + inpaint_model=False, noise_mask_feather=0, scheduler_func=None): + + if noise_mask is not None: + noise_mask = utils.tensor_gaussian_blur_mask(noise_mask, noise_mask_feather) + noise_mask = noise_mask.squeeze(3) + + if noise_mask_feather > 0 and 'denoise_mask_function' not in model.model_options: + model = nodes_differential_diffusion.DifferentialDiffusion().apply(model)[0] + + if wildcard_opt is not None and wildcard_opt != "": + model, _, wildcard_positive = wildcards.process_with_loras(wildcard_opt, model, clip) + + if wildcard_opt_concat_mode == "concat": + positive = nodes.ConditioningConcat().concat(positive, wildcard_positive)[0] + else: + positive = wildcard_positive + positive = [positive[0].copy()] + if 'pooled_output' in wildcard_positive[0][1]: + positive[0][1]['pooled_output'] = wildcard_positive[0][1]['pooled_output'] + elif 'pooled_output' in positive[0][1]: + del positive[0][1]['pooled_output'] + + h = image.shape[1] + w = image.shape[2] + + bbox_h = bbox[3] - bbox[1] + bbox_w = bbox[2] - bbox[0] + + # Skip processing if the detected bbox is already larger than the guide_size + if not force_inpaint and bbox_h >= guide_size and bbox_w >= guide_size: + print(f"Detailer: segment skip (enough big)") + return None, None + + if guide_size_for_bbox: # == "bbox" + # Scale up based on the smaller dimension between width and height. + upscale = guide_size / min(bbox_w, bbox_h) + else: + # for cropped_size + upscale = guide_size / min(w, h) + + new_w = int(w * upscale) + new_h = int(h * upscale) + + # safeguard + if 'aitemplate_keep_loaded' in model.model_options: + max_size = min(4096, max_size) + + if new_w > max_size or new_h > max_size: + upscale *= max_size / max(new_w, new_h) + new_w = int(w * upscale) + new_h = int(h * upscale) + + if not force_inpaint: + if upscale <= 1.0: + print(f"Detailer: segment skip [determined upscale factor={upscale}]") + return None, None + + if new_w == 0 or new_h == 0: + print(f"Detailer: segment skip [zero size={new_w, new_h}]") + return None, None + else: + if upscale <= 1.0 or new_w == 0 or new_h == 0: + print(f"Detailer: force inpaint") + upscale = 1.0 + new_w = w + new_h = h + + if detailer_hook is not None: + new_w, new_h = detailer_hook.touch_scaled_size(new_w, new_h) + + print(f"Detailer: segment upscale for ({bbox_w, bbox_h}) | crop region {w, h} x {upscale} -> {new_w, new_h}") + + # upscale + upscaled_image = tensor_resize(image, new_w, new_h) + + cnet_pils = None + if control_net_wrapper is not None: + positive, negative, cnet_pils = control_net_wrapper.apply(positive, negative, upscaled_image, noise_mask) + model, cnet_pils2 = control_net_wrapper.doit_ipadapter(model) + cnet_pils.extend(cnet_pils2) + + # prepare mask + if noise_mask is not None and inpaint_model: + positive, negative, latent_image = nodes.InpaintModelConditioning().encode(positive, negative, upscaled_image, vae, noise_mask) + else: + latent_image = to_latent_image(upscaled_image, vae) + if noise_mask is not None: + latent_image['noise_mask'] = noise_mask + + if detailer_hook is not None: + latent_image = detailer_hook.post_encode(latent_image) + + refined_latent = latent_image + + # ksampler + for i in range(0, cycle): + if detailer_hook is not None: + if detailer_hook is not None: + detailer_hook.set_steps((i, cycle)) + + refined_latent = detailer_hook.cycle_latent(refined_latent) + + model2, seed2, steps2, cfg2, sampler_name2, scheduler2, positive2, negative2, upscaled_latent2, denoise2 = \ + detailer_hook.pre_ksample(model, seed+i, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise) + noise, is_touched = detailer_hook.get_custom_noise(seed+i, torch.zeros(latent_image['samples'].size()), is_touched=False) + if not is_touched: + noise = None + else: + model2, seed2, steps2, cfg2, sampler_name2, scheduler2, positive2, negative2, upscaled_latent2, denoise2 = \ + model, seed + i, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise + noise = None + + refined_latent = impact_sampling.ksampler_wrapper(model2, seed2, steps2, cfg2, sampler_name2, scheduler2, positive2, negative2, + refined_latent, denoise2, refiner_ratio, refiner_model, refiner_clip, refiner_positive, refiner_negative, + noise=noise, scheduler_func=scheduler_func) + + if detailer_hook is not None: + refined_latent = detailer_hook.pre_decode(refined_latent) + + # non-latent downscale - latent downscale cause bad quality + try: + # try to decode image normally + refined_image = vae.decode(refined_latent['samples']) + except Exception as e: + #usually an out-of-memory exception from the decode, so try a tiled approach + refined_image = vae.decode_tiled(refined_latent["samples"], tile_x=64, tile_y=64, ) + + if detailer_hook is not None: + refined_image = detailer_hook.post_decode(refined_image) + + # downscale + refined_image = tensor_resize(refined_image, w, h) + + # prevent mixing of device + refined_image = refined_image.cpu() + + # don't convert to latent - latent break image + # preserving pil is much better + return refined_image, cnet_pils + + +def enhance_detail_for_animatediff(image_frames, model, clip, vae, guide_size, guide_size_for_bbox, max_size, bbox, seed, steps, cfg, + sampler_name, + scheduler, positive, negative, denoise, noise_mask, + wildcard_opt=None, wildcard_opt_concat_mode=None, + detailer_hook=None, + refiner_ratio=None, refiner_model=None, refiner_clip=None, refiner_positive=None, + refiner_negative=None, control_net_wrapper=None, noise_mask_feather=0, scheduler_func=None): + if noise_mask is not None: + noise_mask = utils.tensor_gaussian_blur_mask(noise_mask, noise_mask_feather) + noise_mask = noise_mask.squeeze(3) + + if noise_mask_feather > 0 and 'denoise_mask_function' not in model.model_options: + model = nodes_differential_diffusion.DifferentialDiffusion().apply(model)[0] + + if wildcard_opt is not None and wildcard_opt != "": + model, _, wildcard_positive = wildcards.process_with_loras(wildcard_opt, model, clip) + + if wildcard_opt_concat_mode == "concat": + positive = nodes.ConditioningConcat().concat(positive, wildcard_positive)[0] + else: + positive = wildcard_positive + + h = image_frames.shape[1] + w = image_frames.shape[2] + + bbox_h = bbox[3] - bbox[1] + bbox_w = bbox[2] - bbox[0] + + # Skip processing if the detected bbox is already larger than the guide_size + if guide_size_for_bbox: # == "bbox" + # Scale up based on the smaller dimension between width and height. + upscale = guide_size / min(bbox_w, bbox_h) + else: + # for cropped_size + upscale = guide_size / min(w, h) + + new_w = int(w * upscale) + new_h = int(h * upscale) + + # safeguard + if 'aitemplate_keep_loaded' in model.model_options: + max_size = min(4096, max_size) + + if new_w > max_size or new_h > max_size: + upscale *= max_size / max(new_w, new_h) + new_w = int(w * upscale) + new_h = int(h * upscale) + + if upscale <= 1.0 or new_w == 0 or new_h == 0: + print(f"Detailer: force inpaint") + upscale = 1.0 + new_w = w + new_h = h + + if detailer_hook is not None: + new_w, new_h = detailer_hook.touch_scaled_size(new_w, new_h) + + print(f"Detailer: segment upscale for ({bbox_w, bbox_h}) | crop region {w, h} x {upscale} -> {new_w, new_h}") + + # upscale the mask tensor by a factor of 2 using bilinear interpolation + if isinstance(noise_mask, np.ndarray): + noise_mask = torch.from_numpy(noise_mask) + + if len(noise_mask.shape) == 2: + noise_mask = noise_mask.unsqueeze(0) + else: # == 3 + noise_mask = noise_mask + + upscaled_mask = None + + for single_mask in noise_mask: + single_mask = single_mask.unsqueeze(0).unsqueeze(0) + upscaled_single_mask = torch.nn.functional.interpolate(single_mask, size=(new_h, new_w), mode='bilinear', align_corners=False) + upscaled_single_mask = upscaled_single_mask.squeeze(0) + + if upscaled_mask is None: + upscaled_mask = upscaled_single_mask + else: + upscaled_mask = torch.cat((upscaled_mask, upscaled_single_mask), dim=0) + + latent_frames = None + for image in image_frames: + image = torch.from_numpy(image).unsqueeze(0) + + # upscale + upscaled_image = tensor_resize(image, new_w, new_h) + + # ksampler + samples = to_latent_image(upscaled_image, vae)['samples'] + + if latent_frames is None: + latent_frames = samples + else: + latent_frames = torch.concat((latent_frames, samples), dim=0) + + cnet_images = None + if control_net_wrapper is not None: + positive, negative, cnet_images = control_net_wrapper.apply(positive, negative, torch.from_numpy(image_frames), noise_mask, use_acn=True) + + if len(upscaled_mask) != len(image_frames) and len(upscaled_mask) > 1: + print(f"[Impact Pack] WARN: DetailerForAnimateDiff - The number of the mask frames({len(upscaled_mask)}) and the image frames({len(image_frames)}) are different. Combine the mask frames and apply.") + combined_mask = upscaled_mask[0].to(torch.uint8) + + for frame_mask in upscaled_mask[1:]: + combined_mask |= (frame_mask * 255).to(torch.uint8) + + combined_mask = (combined_mask/255.0).to(torch.float32) + + upscaled_mask = combined_mask.expand(len(image_frames), -1, -1) + upscaled_mask = utils.to_binary_mask(upscaled_mask, 0.1) + + latent = { + 'noise_mask': upscaled_mask, + 'samples': latent_frames + } + + if detailer_hook is not None: + latent = detailer_hook.post_encode(latent) + + refined_latent = impact_sampling.ksampler_wrapper(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, + latent, denoise, refiner_ratio, refiner_model, refiner_clip, refiner_positive, refiner_negative, scheduler_func=scheduler_func) + + if detailer_hook is not None: + refined_latent = detailer_hook.pre_decode(refined_latent) + + refined_image_frames = None + for refined_sample in refined_latent['samples']: + refined_sample = refined_sample.unsqueeze(0) + + # non-latent downscale - latent downscale cause bad quality + refined_image = vae.decode(refined_sample) + + if refined_image_frames is None: + refined_image_frames = refined_image + else: + refined_image_frames = torch.concat((refined_image_frames, refined_image), dim=0) + + if detailer_hook is not None: + refined_image_frames = detailer_hook.post_decode(refined_image_frames) + + refined_image_frames = nodes.ImageScale().upscale(image=refined_image_frames, upscale_method='lanczos', width=w, height=h, crop='disabled')[0] + + return refined_image_frames, cnet_images + + +def composite_to(dest_latent, crop_region, src_latent): + x1 = crop_region[0] + y1 = crop_region[1] + + # composite to original latent + lc = nodes.LatentComposite() + orig_image = lc.composite(dest_latent, src_latent, x1, y1) + + return orig_image[0] + + +def sam_predict(predictor, points, plabs, bbox, threshold): + point_coords = None if not points else np.array(points) + point_labels = None if not plabs else np.array(plabs) + + box = np.array([bbox]) if bbox is not None else None + + cur_masks, scores, _ = predictor.predict(point_coords=point_coords, point_labels=point_labels, box=box) + + total_masks = [] + + selected = False + max_score = 0 + max_mask = None + for idx in range(len(scores)): + if scores[idx] > max_score: + max_score = scores[idx] + max_mask = cur_masks[idx] + + if scores[idx] >= threshold: + selected = True + total_masks.append(cur_masks[idx]) + else: + pass + + if not selected and max_mask is not None: + total_masks.append(max_mask) + + return total_masks + + +class SAMWrapper: + def __init__(self, model, is_auto_mode, safe_to_gpu=None): + self.model = model + self.safe_to_gpu = safe_to_gpu if safe_to_gpu is not None else SafeToGPU_stub() + self.is_auto_mode = is_auto_mode + + def prepare_device(self): + if self.is_auto_mode: + device = comfy.model_management.get_torch_device() + self.safe_to_gpu.to_device(self.model, device=device) + + def release_device(self): + if self.is_auto_mode: + self.model.to(device="cpu") + + def predict(self, image, points, plabs, bbox, threshold): + predictor = SamPredictor(self.model) + predictor.set_image(image, "RGB") + + return sam_predict(predictor, points, plabs, bbox, threshold) + + +class ESAMWrapper: + def __init__(self, model, device): + self.model = model + self.func_inference = nodes.NODE_CLASS_MAPPINGS['Yoloworld_ESAM_Zho'] + self.device = device + + def prepare_device(self): + pass + + def release_device(self): + pass + + def predict(self, image, points, plabs, bbox, threshold): + if self.device == 'CPU': + self.device = 'cpu' + else: + self.device = 'cuda' + + detected_masks = self.func_inference.inference_sam_with_boxes(image=image, xyxy=[bbox], model=self.model, device=self.device) + return [detected_masks.squeeze(0)] + + +def make_sam_mask(sam, segs, image, detection_hint, dilation, + threshold, bbox_expansion, mask_hint_threshold, mask_hint_use_negative): + + if not hasattr(sam, 'sam_wrapper'): + raise Exception("[Impact Pack] Invalid SAMLoader is connected. Make sure 'SAMLoader (Impact)'.\nKnown issue: The ComfyUI-YOLO node overrides the SAMLoader (Impact), making it unusable. You need to uninstall ComfyUI-YOLO.\n\n\n") + + sam_obj = sam.sam_wrapper + sam_obj.prepare_device() + + try: + image = np.clip(255. * image.cpu().numpy().squeeze(), 0, 255).astype(np.uint8) + + total_masks = [] + + use_small_negative = mask_hint_use_negative == "Small" + + # seg_shape = segs[0] + segs = segs[1] + if detection_hint == "mask-points": + points = [] + plabs = [] + + for i in range(len(segs)): + bbox = segs[i].bbox + center = center_of_bbox(segs[i].bbox) + points.append(center) + + # small point is background, big point is foreground + if use_small_negative and bbox[2] - bbox[0] < 10: + plabs.append(0) + else: + plabs.append(1) + + detected_masks = sam_obj.predict(image, points, plabs, None, threshold) + total_masks += detected_masks + + else: + for i in range(len(segs)): + bbox = segs[i].bbox + center = center_of_bbox(bbox) + + x1 = max(bbox[0] - bbox_expansion, 0) + y1 = max(bbox[1] - bbox_expansion, 0) + x2 = min(bbox[2] + bbox_expansion, image.shape[1]) + y2 = min(bbox[3] + bbox_expansion, image.shape[0]) + + dilated_bbox = [x1, y1, x2, y2] + + points = [] + plabs = [] + if detection_hint == "center-1": + points.append(center) + plabs = [1] # 1 = foreground point, 0 = background point + + elif detection_hint == "horizontal-2": + gap = (x2 - x1) / 3 + points.append((x1 + gap, center[1])) + points.append((x1 + gap * 2, center[1])) + plabs = [1, 1] + + elif detection_hint == "vertical-2": + gap = (y2 - y1) / 3 + points.append((center[0], y1 + gap)) + points.append((center[0], y1 + gap * 2)) + plabs = [1, 1] + + elif detection_hint == "rect-4": + x_gap = (x2 - x1) / 3 + y_gap = (y2 - y1) / 3 + points.append((x1 + x_gap, center[1])) + points.append((x1 + x_gap * 2, center[1])) + points.append((center[0], y1 + y_gap)) + points.append((center[0], y1 + y_gap * 2)) + plabs = [1, 1, 1, 1] + + elif detection_hint == "diamond-4": + x_gap = (x2 - x1) / 3 + y_gap = (y2 - y1) / 3 + points.append((x1 + x_gap, y1 + y_gap)) + points.append((x1 + x_gap * 2, y1 + y_gap)) + points.append((x1 + x_gap, y1 + y_gap * 2)) + points.append((x1 + x_gap * 2, y1 + y_gap * 2)) + plabs = [1, 1, 1, 1] + + elif detection_hint == "mask-point-bbox": + center = center_of_bbox(segs[i].bbox) + points.append(center) + plabs = [1] + + elif detection_hint == "mask-area": + points, plabs = gen_detection_hints_from_mask_area(segs[i].crop_region[0], segs[i].crop_region[1], + segs[i].cropped_mask, + mask_hint_threshold, use_small_negative) + + if mask_hint_use_negative == "Outter": + npoints, nplabs = gen_negative_hints(image.shape[0], image.shape[1], + segs[i].crop_region[0], segs[i].crop_region[1], + segs[i].crop_region[2], segs[i].crop_region[3]) + + points += npoints + plabs += nplabs + + detected_masks = sam_obj.predict(image, points, plabs, dilated_bbox, threshold) + total_masks += detected_masks + + # merge every collected masks + mask = combine_masks2(total_masks) + + finally: + sam_obj.release_device() + + if mask is not None: + mask = mask.float() + mask = dilate_mask(mask.cpu().numpy(), dilation) + mask = torch.from_numpy(mask) + else: + size = image.shape[0], image.shape[1] + mask = torch.zeros(size, dtype=torch.float32, device="cpu") # empty mask + + mask = utils.make_3d_mask(mask) + return mask + + +def generate_detection_hints(image, seg, center, detection_hint, dilated_bbox, mask_hint_threshold, use_small_negative, + mask_hint_use_negative): + [x1, y1, x2, y2] = dilated_bbox + + points = [] + plabs = [] + if detection_hint == "center-1": + points.append(center) + plabs = [1] # 1 = foreground point, 0 = background point + + elif detection_hint == "horizontal-2": + gap = (x2 - x1) / 3 + points.append((x1 + gap, center[1])) + points.append((x1 + gap * 2, center[1])) + plabs = [1, 1] + + elif detection_hint == "vertical-2": + gap = (y2 - y1) / 3 + points.append((center[0], y1 + gap)) + points.append((center[0], y1 + gap * 2)) + plabs = [1, 1] + + elif detection_hint == "rect-4": + x_gap = (x2 - x1) / 3 + y_gap = (y2 - y1) / 3 + points.append((x1 + x_gap, center[1])) + points.append((x1 + x_gap * 2, center[1])) + points.append((center[0], y1 + y_gap)) + points.append((center[0], y1 + y_gap * 2)) + plabs = [1, 1, 1, 1] + + elif detection_hint == "diamond-4": + x_gap = (x2 - x1) / 3 + y_gap = (y2 - y1) / 3 + points.append((x1 + x_gap, y1 + y_gap)) + points.append((x1 + x_gap * 2, y1 + y_gap)) + points.append((x1 + x_gap, y1 + y_gap * 2)) + points.append((x1 + x_gap * 2, y1 + y_gap * 2)) + plabs = [1, 1, 1, 1] + + elif detection_hint == "mask-point-bbox": + center = center_of_bbox(seg.bbox) + points.append(center) + plabs = [1] + + elif detection_hint == "mask-area": + points, plabs = gen_detection_hints_from_mask_area(seg.crop_region[0], seg.crop_region[1], + seg.cropped_mask, + mask_hint_threshold, use_small_negative) + + if mask_hint_use_negative == "Outter": + npoints, nplabs = gen_negative_hints(image.shape[0], image.shape[1], + seg.crop_region[0], seg.crop_region[1], + seg.crop_region[2], seg.crop_region[3]) + + points += npoints + plabs += nplabs + + return points, plabs + + +def convert_and_stack_masks(masks): + if len(masks) == 0: + return None + + mask_tensors = [] + for mask in masks: + mask_array = np.array(mask, dtype=np.uint8) + mask_tensor = torch.from_numpy(mask_array) + mask_tensors.append(mask_tensor) + + stacked_masks = torch.stack(mask_tensors, dim=0) + stacked_masks = stacked_masks.unsqueeze(1) + + return stacked_masks + + +def merge_and_stack_masks(stacked_masks, group_size): + if stacked_masks is None: + return None + + num_masks = stacked_masks.size(0) + merged_masks = [] + + for i in range(0, num_masks, group_size): + subset_masks = stacked_masks[i:i + group_size] + merged_mask = torch.any(subset_masks, dim=0) + merged_masks.append(merged_mask) + + if len(merged_masks) > 0: + merged_masks = torch.stack(merged_masks, dim=0) + + return merged_masks + + +def segs_scale_match(segs, target_shape): + h = segs[0][0] + w = segs[0][1] + + th = target_shape[1] + tw = target_shape[2] + + if (h == th and w == tw) or h == 0 or w == 0: + return segs + + rh = th / h + rw = tw / w + + new_segs = [] + for seg in segs[1]: + cropped_image = seg.cropped_image + cropped_mask = seg.cropped_mask + x1, y1, x2, y2 = seg.crop_region + bx1, by1, bx2, by2 = seg.bbox + + crop_region = int(x1*rw), int(y1*rw), int(x2*rh), int(y2*rh) + bbox = int(bx1*rw), int(by1*rw), int(bx2*rh), int(by2*rh) + new_w = crop_region[2] - crop_region[0] + new_h = crop_region[3] - crop_region[1] + + if isinstance(cropped_mask, np.ndarray): + cropped_mask = torch.from_numpy(cropped_mask) + + if isinstance(cropped_mask, torch.Tensor) and len(cropped_mask.shape) == 3: + cropped_mask = torch.nn.functional.interpolate(cropped_mask.unsqueeze(0), size=(new_h, new_w), mode='bilinear', align_corners=False) + cropped_mask = cropped_mask.squeeze(0) + else: + cropped_mask = torch.nn.functional.interpolate(cropped_mask.unsqueeze(0).unsqueeze(0), size=(new_h, new_w), mode='bilinear', align_corners=False) + cropped_mask = cropped_mask.squeeze(0).squeeze(0).numpy() + + if cropped_image is not None: + cropped_image = tensor_resize(cropped_image if isinstance(cropped_image, torch.Tensor) else torch.from_numpy(cropped_image), new_w, new_h) + cropped_image = cropped_image.numpy() + + new_seg = SEG(cropped_image, cropped_mask, seg.confidence, crop_region, bbox, seg.label, seg.control_net_wrapper) + new_segs.append(new_seg) + + return (th, tw), new_segs + + +# Used Python's slicing feature. stacked_masks[2::3] means starting from index 2, selecting every third tensor with a step size of 3. +# This allows for quickly obtaining the last tensor of every three tensors in stacked_masks. +def every_three_pick_last(stacked_masks): + selected_masks = stacked_masks[2::3] + return selected_masks + + +def make_sam_mask_segmented(sam, segs, image, detection_hint, dilation, + threshold, bbox_expansion, mask_hint_threshold, mask_hint_use_negative): + + if not hasattr(sam, 'sam_wrapper'): + raise Exception("[Impact Pack] Invalid SAMLoader is connected. Make sure 'SAMLoader (Impact)'.") + + sam_obj = sam.sam_wrapper + sam_obj.prepare_device() + + try: + image = np.clip(255. * image.cpu().numpy().squeeze(), 0, 255).astype(np.uint8) + + total_masks = [] + + use_small_negative = mask_hint_use_negative == "Small" + + # seg_shape = segs[0] + segs = segs[1] + if detection_hint == "mask-points": + points = [] + plabs = [] + + for i in range(len(segs)): + bbox = segs[i].bbox + center = center_of_bbox(bbox) + points.append(center) + + # small point is background, big point is foreground + if use_small_negative and bbox[2] - bbox[0] < 10: + plabs.append(0) + else: + plabs.append(1) + + detected_masks = sam_obj.predict(image, points, plabs, None, threshold) + total_masks += detected_masks + + else: + for i in range(len(segs)): + bbox = segs[i].bbox + center = center_of_bbox(bbox) + x1 = max(bbox[0] - bbox_expansion, 0) + y1 = max(bbox[1] - bbox_expansion, 0) + x2 = min(bbox[2] + bbox_expansion, image.shape[1]) + y2 = min(bbox[3] + bbox_expansion, image.shape[0]) + + dilated_bbox = [x1, y1, x2, y2] + + points, plabs = generate_detection_hints(image, segs[i], center, detection_hint, dilated_bbox, + mask_hint_threshold, use_small_negative, + mask_hint_use_negative) + + detected_masks = sam_obj.predict(image, points, plabs, dilated_bbox, threshold) + + total_masks += detected_masks + + # merge every collected masks + mask = combine_masks2(total_masks) + + finally: + sam_obj.release_device() + + mask_working_device = torch.device("cpu") + + if mask is not None: + mask = mask.float() + mask = dilate_mask(mask.cpu().numpy(), dilation) + mask = torch.from_numpy(mask) + mask = mask.to(device=mask_working_device) + else: + # Extracting batch, height and width + height, width, _ = image.shape + mask = torch.zeros( + (height, width), dtype=torch.float32, device=mask_working_device + ) # empty mask + + stacked_masks = convert_and_stack_masks(total_masks) + + return (mask, merge_and_stack_masks(stacked_masks, group_size=3)) + # return every_three_pick_last(stacked_masks) + + +def segs_bitwise_and_mask(segs, mask): + mask = make_2d_mask(mask) + + if mask is None: + print("[SegsBitwiseAndMask] Cannot operate: MASK is empty.") + return ([],) + + items = [] + + mask = (mask.cpu().numpy() * 255).astype(np.uint8) + + for seg in segs[1]: + cropped_mask = (seg.cropped_mask * 255).astype(np.uint8) + crop_region = seg.crop_region + + cropped_mask2 = mask[crop_region[1]:crop_region[3], crop_region[0]:crop_region[2]] + + new_mask = np.bitwise_and(cropped_mask.astype(np.uint8), cropped_mask2) + new_mask = new_mask.astype(np.float32) / 255.0 + + item = SEG(seg.cropped_image, new_mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, None) + items.append(item) + + return segs[0], items + + +def segs_bitwise_subtract_mask(segs, mask): + mask = make_2d_mask(mask) + + if mask is None: + print("[SegsBitwiseSubtractMask] Cannot operate: MASK is empty.") + return ([],) + + items = [] + + mask = (mask.cpu().numpy() * 255).astype(np.uint8) + + for seg in segs[1]: + cropped_mask = (seg.cropped_mask * 255).astype(np.uint8) + crop_region = seg.crop_region + + cropped_mask2 = mask[crop_region[1]:crop_region[3], crop_region[0]:crop_region[2]] + + new_mask = cv2.subtract(cropped_mask.astype(np.uint8), cropped_mask2) + new_mask = new_mask.astype(np.float32) / 255.0 + + item = SEG(seg.cropped_image, new_mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, None) + items.append(item) + + return segs[0], items + + +def apply_mask_to_each_seg(segs, masks): + if masks is None: + print("[SegsBitwiseAndMask] Cannot operate: MASK is empty.") + return (segs[0], [],) + + items = [] + + masks = masks.squeeze(1) + + for seg, mask in zip(segs[1], masks): + cropped_mask = (seg.cropped_mask * 255).astype(np.uint8) + crop_region = seg.crop_region + + cropped_mask2 = (mask.cpu().numpy() * 255).astype(np.uint8) + cropped_mask2 = cropped_mask2[crop_region[1]:crop_region[3], crop_region[0]:crop_region[2]] + + new_mask = np.bitwise_and(cropped_mask.astype(np.uint8), cropped_mask2) + new_mask = new_mask.astype(np.float32) / 255.0 + + item = SEG(seg.cropped_image, new_mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, None) + items.append(item) + + return segs[0], items + + +def dilate_segs(segs, factor): + if factor == 0: + return segs + + new_segs = [] + for seg in segs[1]: + new_mask = dilate_mask(seg.cropped_mask, factor) + new_seg = SEG(seg.cropped_image, new_mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, seg.control_net_wrapper) + new_segs.append(new_seg) + + return (segs[0], new_segs) + + +class ONNXDetector: + onnx_model = None + + def __init__(self, onnx_model): + self.onnx_model = onnx_model + + def detect(self, image, threshold, dilation, crop_factor, drop_size=1, detailer_hook=None): + drop_size = max(drop_size, 1) + try: + import impact.onnx as onnx + + h = image.shape[1] + w = image.shape[2] + + labels, scores, boxes = onnx.onnx_inference(image, self.onnx_model) + + # collect feasible item + result = [] + + for i in range(len(labels)): + if scores[i] > threshold: + item_bbox = boxes[i] + x1, y1, x2, y2 = item_bbox + + if x2 - x1 > drop_size and y2 - y1 > drop_size: # minimum dimension must be (2,2) to avoid squeeze issue + crop_region = make_crop_region(w, h, item_bbox, crop_factor) + + if detailer_hook is not None: + crop_region = item_bbox.post_crop_region(w, h, item_bbox, crop_region) + + crop_x1, crop_y1, crop_x2, crop_y2, = crop_region + + # prepare cropped mask + cropped_mask = np.zeros((crop_y2 - crop_y1, crop_x2 - crop_x1)) + cropped_mask[y1 - crop_y1:y2 - crop_y1, x1 - crop_x1:x2 - crop_x1] = 1 + cropped_mask = dilate_mask(cropped_mask, dilation) + + # make items. just convert the integer label to a string + item = SEG(None, cropped_mask, scores[i], crop_region, item_bbox, str(labels[i]), None) + result.append(item) + + shape = h, w + segs = shape, result + + if detailer_hook is not None and hasattr(detailer_hook, "post_detection"): + segs = detailer_hook.post_detection(segs) + + return segs + except Exception as e: + print(f"ONNXDetector: unable to execute.\n{e}") + pass + + def detect_combined(self, image, threshold, dilation): + return segs_to_combined_mask(self.detect(image, threshold, dilation, 1)) + + def setAux(self, x): + pass + + +def batch_mask_to_segs(mask, combined, crop_factor, bbox_fill, drop_size=1, label='A', crop_min_size=None, detailer_hook=None): + combined_mask = mask.max(dim=0).values + + segs = mask_to_segs(combined_mask, combined, crop_factor, bbox_fill, drop_size, label, crop_min_size, detailer_hook) + + new_segs = [] + for seg in segs[1]: + x1, y1, x2, y2 = seg.crop_region + cropped_mask = mask[:, y1:y2, x1:x2] + item = SEG(None, cropped_mask, 1.0, seg.crop_region, seg.bbox, label, None) + new_segs.append(item) + + return segs[0], new_segs + + +def mask_to_segs(mask, combined, crop_factor, bbox_fill, drop_size=1, label='A', crop_min_size=None, detailer_hook=None, is_contour=True): + drop_size = max(drop_size, 1) + if mask is None: + print("[mask_to_segs] Cannot operate: MASK is empty.") + return ([],) + + if isinstance(mask, np.ndarray): + pass # `mask` is already a NumPy array + else: + try: + mask = mask.numpy() + except AttributeError: + print("[mask_to_segs] Cannot operate: MASK is not a NumPy array or Tensor.") + return ([],) + + if mask is None: + print("[mask_to_segs] Cannot operate: MASK is empty.") + return ([],) + + result = [] + + if len(mask.shape) == 2: + mask = np.expand_dims(mask, axis=0) + + for i in range(mask.shape[0]): + mask_i = mask[i] + + if combined: + indices = np.nonzero(mask_i) + if len(indices[0]) > 0 and len(indices[1]) > 0: + bbox = ( + np.min(indices[1]), + np.min(indices[0]), + np.max(indices[1]), + np.max(indices[0]), + ) + crop_region = make_crop_region( + mask_i.shape[1], mask_i.shape[0], bbox, crop_factor + ) + x1, y1, x2, y2 = crop_region + + if detailer_hook is not None: + crop_region = detailer_hook.post_crop_region(mask_i.shape[1], mask_i.shape[0], bbox, crop_region) + + if x2 - x1 > 0 and y2 - y1 > 0: + cropped_mask = mask_i[y1:y2, x1:x2] + + if bbox_fill: + bx1, by1, bx2, by2 = bbox + cropped_mask = cropped_mask.copy() + cropped_mask[by1:by2, bx1:bx2] = 1.0 + + if cropped_mask is not None: + item = SEG(None, cropped_mask, 1.0, crop_region, bbox, label, None) + result.append(item) + + else: + mask_i_uint8 = (mask_i * 255.0).astype(np.uint8) + contours, ctree = cv2.findContours(mask_i_uint8, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE) + for j, contour in enumerate(contours): + hierarchy = ctree[0][j] + if hierarchy[3] != -1: + continue + + separated_mask = np.zeros_like(mask_i_uint8) + cv2.drawContours(separated_mask, [contour], 0, 255, -1) + separated_mask = np.array(separated_mask / 255.0).astype(np.float32) + + x, y, w, h = cv2.boundingRect(contour) + bbox = x, y, x + w, y + h + crop_region = make_crop_region( + mask_i.shape[1], mask_i.shape[0], bbox, crop_factor, crop_min_size + ) + + if detailer_hook is not None: + crop_region = detailer_hook.post_crop_region(mask_i.shape[1], mask_i.shape[0], bbox, crop_region) + + if w > drop_size and h > drop_size: + if is_contour: + mask_src = separated_mask + else: + mask_src = mask_i * separated_mask + + cropped_mask = np.array( + mask_src[ + crop_region[1]: crop_region[3], + crop_region[0]: crop_region[2], + ] + ) + + if bbox_fill: + cx1, cy1, _, _ = crop_region + bx1 = x - cx1 + bx2 = x+w - cx1 + by1 = y - cy1 + by2 = y+h - cy1 + cropped_mask[by1:by2, bx1:bx2] = 1.0 + + if cropped_mask is not None: + cropped_mask = torch.clip(torch.from_numpy(cropped_mask), 0, 1.0) + item = SEG(None, cropped_mask.numpy(), 1.0, crop_region, bbox, label, None) + result.append(item) + + if not result: + print(f"[mask_to_segs] Empty mask.") + + print(f"# of Detected SEGS: {len(result)}") + # for r in result: + # print(f"\tbbox={r.bbox}, crop={r.crop_region}, label={r.label}") + + # shape: (b,h,w) -> (h,w) + return (mask.shape[1], mask.shape[2]), result + + +def mediapipe_facemesh_to_segs(image, crop_factor, bbox_fill, crop_min_size, drop_size, dilation, face, mouth, left_eyebrow, left_eye, left_pupil, right_eyebrow, right_eye, right_pupil): + parts = { + "face": np.array([0x0A, 0xC8, 0x0A]), + "mouth": np.array([0x0A, 0xB4, 0x0A]), + "left_eyebrow": np.array([0xB4, 0xDC, 0x0A]), + "left_eye": np.array([0xB4, 0xC8, 0x0A]), + "left_pupil": np.array([0xFA, 0xC8, 0x0A]), + "right_eyebrow": np.array([0x0A, 0xDC, 0xB4]), + "right_eye": np.array([0x0A, 0xC8, 0xB4]), + "right_pupil": np.array([0x0A, 0xC8, 0xFA]), + } + + def create_segments(image, color): + image = (image * 255).to(torch.uint8) + image = image.squeeze(0).numpy() + mask = cv2.inRange(image, color, color) + + contours, ctree = cv2.findContours(mask, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE) + mask_list = [] + for i, contour in enumerate(contours): + hierarchy = ctree[0][i] + if hierarchy[3] == -1: + convex_hull = cv2.convexHull(contour) + convex_segment = np.zeros_like(image) + cv2.fillPoly(convex_segment, [convex_hull], (255, 255, 255)) + + convex_segment = np.expand_dims(convex_segment, axis=0).astype(np.float32) / 255.0 + tensor = torch.from_numpy(convex_segment) + mask_tensor = torch.any(tensor != 0, dim=-1).float() + mask_tensor = mask_tensor.squeeze(0) + mask_tensor = torch.from_numpy(dilate_mask(mask_tensor.numpy(), dilation)) + mask_list.append(mask_tensor.unsqueeze(0)) + + return mask_list + + segs = [] + + def create_seg(label): + mask_list = create_segments(image, parts[label]) + for mask in mask_list: + seg = mask_to_segs(mask, False, crop_factor, bbox_fill, drop_size=drop_size, label=label, crop_min_size=crop_min_size) + if len(seg[1]) > 0: + segs.extend(seg[1]) + + if face: + create_seg('face') + + if mouth: + create_seg('mouth') + + if left_eyebrow: + create_seg('left_eyebrow') + + if left_eye: + create_seg('left_eye') + + if left_pupil: + create_seg('left_pupil') + + if right_eyebrow: + create_seg('right_eyebrow') + + if right_eye: + create_seg('right_eye') + + if right_pupil: + create_seg('right_pupil') + + return (image.shape[1], image.shape[2]), segs + + +def segs_to_combined_mask(segs): + shape = segs[0] + h = shape[0] + w = shape[1] + + mask = np.zeros((h, w), dtype=np.uint8) + + for seg in segs[1]: + cropped_mask = seg.cropped_mask + crop_region = seg.crop_region + mask[crop_region[1]:crop_region[3], crop_region[0]:crop_region[2]] |= (cropped_mask * 255).astype(np.uint8) + + return torch.from_numpy(mask.astype(np.float32) / 255.0) + + +def segs_to_masklist(segs): + shape = segs[0] + h = shape[0] + w = shape[1] + + masks = [] + for seg in segs[1]: + if isinstance(seg.cropped_mask, np.ndarray): + cropped_mask = torch.from_numpy(seg.cropped_mask) + else: + cropped_mask = seg.cropped_mask + + if cropped_mask.ndim == 2: + cropped_mask = cropped_mask.unsqueeze(0) + + n = len(cropped_mask) + + mask = torch.zeros((n, h, w), dtype=torch.uint8) + crop_region = seg.crop_region + mask[:, crop_region[1]:crop_region[3], crop_region[0]:crop_region[2]] |= (cropped_mask * 255).to(torch.uint8) + mask = (mask / 255.0).to(torch.float32) + + for x in mask: + masks.append(x) + + if len(masks) == 0: + empty_mask = torch.zeros((h, w), dtype=torch.float32, device="cpu") + masks = [empty_mask] + + return masks + + +def vae_decode(vae, samples, use_tile, hook, tile_size=512): + if use_tile: + pixels = nodes.VAEDecodeTiled().decode(vae, samples, tile_size)[0] + else: + pixels = nodes.VAEDecode().decode(vae, samples)[0] + + if hook is not None: + pixels = hook.post_decode(pixels) + + return pixels + + +def vae_encode(vae, pixels, use_tile, hook, tile_size=512): + if use_tile: + samples = nodes.VAEEncodeTiled().encode(vae, pixels, tile_size)[0] + else: + samples = nodes.VAEEncode().encode(vae, pixels)[0] + + if hook is not None: + samples = hook.post_encode(samples) + + return samples + + +def latent_upscale_on_pixel_space_shape(samples, scale_method, w, h, vae, use_tile=False, tile_size=512, save_temp_prefix=None, hook=None): + return latent_upscale_on_pixel_space_shape2(samples, scale_method, w, h, vae, use_tile, tile_size, save_temp_prefix, hook)[0] + + +def latent_upscale_on_pixel_space_shape2(samples, scale_method, w, h, vae, use_tile=False, tile_size=512, save_temp_prefix=None, hook=None): + pixels = vae_decode(vae, samples, use_tile, hook, tile_size=tile_size) + + if save_temp_prefix is not None: + nodes.PreviewImage().save_images(pixels, filename_prefix=save_temp_prefix) + + pixels = nodes.ImageScale().upscale(pixels, scale_method, int(w), int(h), False)[0] + + old_pixels = pixels + if hook is not None: + pixels = hook.post_upscale(pixels) + + return (vae_encode(vae, pixels, use_tile, hook, tile_size=tile_size), old_pixels) + + +def latent_upscale_on_pixel_space(samples, scale_method, scale_factor, vae, use_tile=False, tile_size=512, save_temp_prefix=None, hook=None): + return latent_upscale_on_pixel_space2(samples, scale_method, scale_factor, vae, use_tile, tile_size, save_temp_prefix, hook)[0] + + +def latent_upscale_on_pixel_space2(samples, scale_method, scale_factor, vae, use_tile=False, tile_size=512, save_temp_prefix=None, hook=None): + pixels = vae_decode(vae, samples, use_tile, hook, tile_size=tile_size) + + if save_temp_prefix is not None: + nodes.PreviewImage().save_images(pixels, filename_prefix=save_temp_prefix) + + w = pixels.shape[2] * scale_factor + h = pixels.shape[1] * scale_factor + pixels = nodes.ImageScale().upscale(pixels, scale_method, int(w), int(h), False)[0] + + old_pixels = pixels + if hook is not None: + pixels = hook.post_upscale(pixels) + + return (vae_encode(vae, pixels, use_tile, hook, tile_size=tile_size), old_pixels) + + +def latent_upscale_on_pixel_space_with_model_shape(samples, scale_method, upscale_model, new_w, new_h, vae, use_tile=False, tile_size=512, save_temp_prefix=None, hook=None): + return latent_upscale_on_pixel_space_with_model_shape2(samples, scale_method, upscale_model, new_w, new_h, vae, use_tile, tile_size, save_temp_prefix, hook)[0] + + +def latent_upscale_on_pixel_space_with_model_shape2(samples, scale_method, upscale_model, new_w, new_h, vae, use_tile=False, tile_size=512, save_temp_prefix=None, hook=None): + pixels = vae_decode(vae, samples, use_tile, hook, tile_size=tile_size) + + if save_temp_prefix is not None: + nodes.PreviewImage().save_images(pixels, filename_prefix=save_temp_prefix) + + w = pixels.shape[2] + + # upscale by model upscaler + current_w = w + while current_w < new_w: + pixels = model_upscale.ImageUpscaleWithModel().upscale(upscale_model, pixels)[0] + current_w = pixels.shape[2] + if current_w == w: + print(f"[latent_upscale_on_pixel_space_with_model] x1 upscale model selected") + break + + # downscale to target scale + pixels = nodes.ImageScale().upscale(pixels, scale_method, int(new_w), int(new_h), False)[0] + + old_pixels = pixels + if hook is not None: + pixels = hook.post_upscale(pixels) + + return (vae_encode(vae, pixels, use_tile, hook, tile_size=tile_size), old_pixels) + + +def latent_upscale_on_pixel_space_with_model(samples, scale_method, upscale_model, scale_factor, vae, use_tile=False, + tile_size=512, save_temp_prefix=None, hook=None): + return latent_upscale_on_pixel_space_with_model2(samples, scale_method, upscale_model, scale_factor, vae, use_tile, tile_size, save_temp_prefix, hook)[0] + +def latent_upscale_on_pixel_space_with_model2(samples, scale_method, upscale_model, scale_factor, vae, use_tile=False, + tile_size=512, save_temp_prefix=None, hook=None): + pixels = vae_decode(vae, samples, use_tile, hook, tile_size=tile_size) + + if save_temp_prefix is not None: + nodes.PreviewImage().save_images(pixels, filename_prefix=save_temp_prefix) + + w = pixels.shape[2] + h = pixels.shape[1] + + new_w = w * scale_factor + new_h = h * scale_factor + + # upscale by model upscaler + current_w = w + while current_w < new_w: + pixels = model_upscale.ImageUpscaleWithModel().upscale(upscale_model, pixels)[0] + current_w = pixels.shape[2] + if current_w == w: + print(f"[latent_upscale_on_pixel_space_with_model] x1 upscale model selected") + break + + # downscale to target scale + pixels = nodes.ImageScale().upscale(pixels, scale_method, int(new_w), int(new_h), False)[0] + + old_pixels = pixels + if hook is not None: + pixels = hook.post_upscale(pixels) + + return (vae_encode(vae, pixels, use_tile, hook, tile_size=tile_size), old_pixels) + + +class TwoSamplersForMaskUpscaler: + def __init__(self, scale_method, sample_schedule, use_tiled_vae, base_sampler, mask_sampler, mask, vae, + full_sampler_opt=None, upscale_model_opt=None, hook_base_opt=None, hook_mask_opt=None, + hook_full_opt=None, + tile_size=512): + + mask = make_2d_mask(mask) + + mask = mask.reshape((-1, 1, mask.shape[-2], mask.shape[-1])) + + self.params = scale_method, sample_schedule, use_tiled_vae, base_sampler, mask_sampler, mask, vae + self.upscale_model = upscale_model_opt + self.full_sampler = full_sampler_opt + self.hook_base = hook_base_opt + self.hook_mask = hook_mask_opt + self.hook_full = hook_full_opt + self.use_tiled_vae = use_tiled_vae + self.tile_size = tile_size + self.is_tiled = False + self.vae = vae + + def upscale(self, step_info, samples, upscale_factor, save_temp_prefix=None): + scale_method, sample_schedule, use_tiled_vae, base_sampler, mask_sampler, mask, vae = self.params + + mask = make_2d_mask(mask) + + self.prepare_hook(step_info) + + # upscale latent + if self.upscale_model is None: + upscaled_latent = latent_upscale_on_pixel_space(samples, scale_method, upscale_factor, vae, + use_tile=self.use_tiled_vae, + save_temp_prefix=save_temp_prefix, + hook=self.hook_base, tile_size=self.tile_size) + else: + upscaled_latent = latent_upscale_on_pixel_space_with_model(samples, scale_method, self.upscale_model, + upscale_factor, vae, + use_tile=self.use_tiled_vae, + save_temp_prefix=save_temp_prefix, + hook=self.hook_mask, tile_size=self.tile_size) + + return self.do_samples(step_info, base_sampler, mask_sampler, sample_schedule, mask, upscaled_latent) + + def prepare_hook(self, step_info): + if self.hook_base is not None: + self.hook_base.set_steps(step_info) + if self.hook_mask is not None: + self.hook_mask.set_steps(step_info) + if self.hook_full is not None: + self.hook_full.set_steps(step_info) + + def upscale_shape(self, step_info, samples, w, h, save_temp_prefix=None): + scale_method, sample_schedule, use_tiled_vae, base_sampler, mask_sampler, mask, vae = self.params + + mask = make_2d_mask(mask) + + self.prepare_hook(step_info) + + # upscale latent + if self.upscale_model is None: + upscaled_latent = latent_upscale_on_pixel_space_shape(samples, scale_method, w, h, vae, + use_tile=self.use_tiled_vae, + save_temp_prefix=save_temp_prefix, + hook=self.hook_base, + tile_size=self.tile_size) + else: + upscaled_latent = latent_upscale_on_pixel_space_with_model_shape(samples, scale_method, self.upscale_model, + w, h, vae, + use_tile=self.use_tiled_vae, + save_temp_prefix=save_temp_prefix, + hook=self.hook_mask, + tile_size=self.tile_size) + + return self.do_samples(step_info, base_sampler, mask_sampler, sample_schedule, mask, upscaled_latent) + + def is_full_sample_time(self, step_info, sample_schedule): + cur_step, total_step = step_info + + # make start from 1 instead of zero + cur_step += 1 + total_step += 1 + + if sample_schedule == "none": + return False + + elif sample_schedule == "interleave1": + return cur_step % 2 == 0 + + elif sample_schedule == "interleave2": + return cur_step % 3 == 0 + + elif sample_schedule == "interleave3": + return cur_step % 4 == 0 + + elif sample_schedule == "last1": + return cur_step == total_step + + elif sample_schedule == "last2": + return cur_step >= total_step - 1 + + elif sample_schedule == "interleave1+last1": + return cur_step % 2 == 0 or cur_step >= total_step - 1 + + elif sample_schedule == "interleave2+last1": + return cur_step % 2 == 0 or cur_step >= total_step - 1 + + elif sample_schedule == "interleave3+last1": + return cur_step % 2 == 0 or cur_step >= total_step - 1 + + def do_samples(self, step_info, base_sampler, mask_sampler, sample_schedule, mask, upscaled_latent): + mask = make_2d_mask(mask) + + if self.is_full_sample_time(step_info, sample_schedule): + print(f"step_info={step_info} / full time") + + upscaled_latent = base_sampler.sample(upscaled_latent, self.hook_base) + sampler = self.full_sampler if self.full_sampler is not None else base_sampler + return sampler.sample(upscaled_latent, self.hook_full) + + else: + print(f"step_info={step_info} / non-full time") + # upscale mask + if mask.ndim == 2: + mask = mask[None, :, :, None] + upscaled_mask = F.interpolate(mask, size=(upscaled_latent['samples'].shape[2], upscaled_latent['samples'].shape[3]), mode='bilinear', align_corners=True) + upscaled_mask = upscaled_mask[:, :, :upscaled_latent['samples'].shape[2], :upscaled_latent['samples'].shape[3]] + + # base sampler + upscaled_inv_mask = torch.where(upscaled_mask != 1.0, torch.tensor(1.0), torch.tensor(0.0)) + upscaled_latent['noise_mask'] = upscaled_inv_mask + upscaled_latent = base_sampler.sample(upscaled_latent, self.hook_base) + + # mask sampler + upscaled_latent['noise_mask'] = upscaled_mask + upscaled_latent = mask_sampler.sample(upscaled_latent, self.hook_mask) + + # remove mask + del upscaled_latent['noise_mask'] + return upscaled_latent + + +class PixelKSampleUpscaler: + def __init__(self, scale_method, model, vae, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise, + use_tiled_vae, upscale_model_opt=None, hook_opt=None, tile_size=512, scheduler_func=None, + tile_cnet_opt=None, tile_cnet_strength=1.0): + self.params = scale_method, model, vae, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise + self.upscale_model = upscale_model_opt + self.hook = hook_opt + self.use_tiled_vae = use_tiled_vae + self.tile_size = tile_size + self.is_tiled = False + self.vae = vae + self.scheduler_func = scheduler_func + self.tile_cnet = tile_cnet_opt + self.tile_cnet_strength = tile_cnet_strength + + def sample(self, model, seed, steps, cfg, sampler_name, scheduler, positive, negative, upscaled_latent, denoise, images): + if self.tile_cnet is not None: + image_batch, image_w, image_h, _ = images.shape + if image_batch > 1: + warnings.warn('Multiple latents in batch, Tile ControlNet being ignored') + else: + if 'TilePreprocessor' not in nodes.NODE_CLASS_MAPPINGS: + raise RuntimeError("'TilePreprocessor' node (from comfyui_controlnet_aux) isn't installed.") + preprocessor = nodes.NODE_CLASS_MAPPINGS['TilePreprocessor']() + # might add capacity to set pyrUp_iters later, not needed for now though + preprocessed = preprocessor.execute(images, pyrUp_iters=3, resolution=min(image_w, image_h))[0] + apply_cnet = getattr(nodes.ControlNetApply(), nodes.ControlNetApply.FUNCTION) + positive = apply_cnet(positive, self.tile_cnet, preprocessed, strength=self.tile_cnet_strength)[0] + + refined_latent = impact_sampling.impact_sample(model, seed, steps, cfg, sampler_name, scheduler, + positive, negative, upscaled_latent, denoise, scheduler_func=self.scheduler_func) + + return refined_latent + + def upscale(self, step_info, samples, upscale_factor, save_temp_prefix=None): + scale_method, model, vae, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise = self.params + + if self.hook is not None: + self.hook.set_steps(step_info) + + if self.upscale_model is None: + upscaled_latent, upscaled_images = \ + latent_upscale_on_pixel_space2(samples, scale_method, upscale_factor, vae, + use_tile=self.use_tiled_vae, + save_temp_prefix=save_temp_prefix, hook=self.hook, tile_size=512) + else: + upscaled_latent, upscaled_images = \ + latent_upscale_on_pixel_space_with_model2(samples, scale_method, self.upscale_model, + upscale_factor, vae, + use_tile=self.use_tiled_vae, + save_temp_prefix=save_temp_prefix, + hook=self.hook, + tile_size=self.tile_size) + + if self.hook is not None: + model, seed, steps, cfg, sampler_name, scheduler, positive, negative, upscaled_latent, denoise = \ + self.hook.pre_ksample(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, + upscaled_latent, denoise) + + if 'noise_mask' in samples: + upscaled_latent['noise_mask'] = samples['noise_mask'] + + refined_latent = self.sample(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, upscaled_latent, denoise, upscaled_images) + return refined_latent + + def upscale_shape(self, step_info, samples, w, h, save_temp_prefix=None): + scale_method, model, vae, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise = self.params + + if self.hook is not None: + self.hook.set_steps(step_info) + + if self.upscale_model is None: + upscaled_latent, upscaled_images = \ + latent_upscale_on_pixel_space_shape2(samples, scale_method, w, h, vae, + use_tile=self.use_tiled_vae, + save_temp_prefix=save_temp_prefix, hook=self.hook, + tile_size=self.tile_size) + else: + upscaled_latent, upscaled_images = \ + latent_upscale_on_pixel_space_with_model_shape2(samples, scale_method, self.upscale_model, + w, h, vae, + use_tile=self.use_tiled_vae, + save_temp_prefix=save_temp_prefix, + hook=self.hook, + tile_size=self.tile_size) + + if self.hook is not None: + model, seed, steps, cfg, sampler_name, scheduler, positive, negative, upscaled_latent, denoise = \ + self.hook.pre_ksample(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, + upscaled_latent, denoise) + + if 'noise_mask' in samples: + upscaled_latent['noise_mask'] = samples['noise_mask'] + + refined_latent = self.sample(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, upscaled_latent, denoise, upscaled_images) + return refined_latent + + +class IPAdapterWrapper: + def __init__(self, ipadapter_pipe, weight, noise, weight_type, start_at, end_at, unfold_batch, weight_v2, reference_image, neg_image=None, prev_control_net=None, combine_embeds='concat'): + self.reference_image = reference_image + self.ipadapter_pipe = ipadapter_pipe + self.weight = weight + self.weight_type = weight_type + self.noise = noise + self.start_at = start_at + self.end_at = end_at + self.unfold_batch = unfold_batch + self.prev_control_net = prev_control_net + self.weight_v2 = weight_v2 + self.image = reference_image + self.neg_image = neg_image + self.combine_embeds = combine_embeds + + # name 'apply_ipadapter' isn't allowed + def doit_ipadapter(self, model): + cnet_image_list = [self.image] + prev_cnet_images = [] + + if 'IPAdapterAdvanced' not in nodes.NODE_CLASS_MAPPINGS: + if 'IPAdapterApply' in nodes.NODE_CLASS_MAPPINGS: + raise Exception(f"[ERROR] 'ComfyUI IPAdapter Plus' is outdated.") + + utils.try_install_custom_node('https://github.com/cubiq/ComfyUI_IPAdapter_plus', + "To use 'IPAdapterApplySEGS' node, 'ComfyUI IPAdapter Plus' extension is required.") + raise Exception(f"[ERROR] To use IPAdapterApplySEGS, you need to install 'ComfyUI IPAdapter Plus'") + + obj = nodes.NODE_CLASS_MAPPINGS['IPAdapterAdvanced'] + + ipadapter, _, clip_vision, insightface, lora_loader = self.ipadapter_pipe + model = lora_loader(model) + + if self.prev_control_net is not None: + model, prev_cnet_images = self.prev_control_net.doit_ipadapter(model) + + model = obj().apply_ipadapter(model=model, ipadapter=ipadapter, weight=self.weight, weight_type=self.weight_type, + start_at=self.start_at, end_at=self.end_at, combine_embeds=self.combine_embeds, + clip_vision=clip_vision, image=self.image, image_negative=self.neg_image, attn_mask=None, + insightface=insightface, weight_faceidv2=self.weight_v2)[0] + + cnet_image_list.extend(prev_cnet_images) + + return model, cnet_image_list + + def apply(self, positive, negative, image, mask=None, use_acn=False): + if self.prev_control_net is not None: + return self.prev_control_net.apply(positive, negative, image, mask, use_acn=use_acn) + else: + return positive, negative, [] + + +class ControlNetWrapper: + def __init__(self, control_net, strength, preprocessor, prev_control_net=None, original_size=None, crop_region=None, control_image=None): + self.control_net = control_net + self.strength = strength + self.preprocessor = preprocessor + self.prev_control_net = prev_control_net + + if original_size is not None and crop_region is not None and control_image is not None: + self.control_image = utils.tensor_resize(control_image, original_size[1], original_size[0]) + self.control_image = torch.tensor(utils.tensor_crop(self.control_image, crop_region)) + else: + self.control_image = None + + def apply(self, positive, negative, image, mask=None, use_acn=False): + cnet_image_list = [] + prev_cnet_images = [] + + if self.prev_control_net is not None: + positive, negative, prev_cnet_images = self.prev_control_net.apply(positive, negative, image, mask, use_acn=use_acn) + + if self.control_image is not None: + cnet_image = self.control_image + elif self.preprocessor is not None: + cnet_image = self.preprocessor.apply(image, mask) + else: + cnet_image = image + + cnet_image_list.extend(prev_cnet_images) + cnet_image_list.append(cnet_image) + + if use_acn: + if "ACN_AdvancedControlNetApply" in nodes.NODE_CLASS_MAPPINGS: + acn = nodes.NODE_CLASS_MAPPINGS['ACN_AdvancedControlNetApply']() + positive, negative, _ = acn.apply_controlnet(positive=positive, negative=negative, control_net=self.control_net, image=cnet_image, + strength=self.strength, start_percent=0.0, end_percent=1.0) + else: + utils.try_install_custom_node('https://github.com/BlenderNeko/ComfyUI_TiledKSampler', + "To use 'ControlNetWrapper' for AnimateDiff, 'ComfyUI-Advanced-ControlNet' extension is required.") + raise Exception("'ACN_AdvancedControlNetApply' node isn't installed.") + else: + positive = nodes.ControlNetApply().apply_controlnet(positive, self.control_net, cnet_image, self.strength)[0] + + return positive, negative, cnet_image_list + + def doit_ipadapter(self, model): + if self.prev_control_net is not None: + return self.prev_control_net.doit_ipadapter(model) + else: + return model, [] + + +class ControlNetAdvancedWrapper: + def __init__(self, control_net, strength, start_percent, end_percent, preprocessor, prev_control_net=None, + original_size=None, crop_region=None, control_image=None, vae=None): + self.control_net = control_net + self.strength = strength + self.preprocessor = preprocessor + self.prev_control_net = prev_control_net + self.start_percent = start_percent + self.end_percent = end_percent + self.vae = vae + + if original_size is not None and crop_region is not None and control_image is not None: + self.control_image = utils.tensor_resize(control_image, original_size[1], original_size[0]) + self.control_image = torch.tensor(utils.tensor_crop(self.control_image, crop_region)) + else: + self.control_image = None + + def doit_ipadapter(self, model): + if self.prev_control_net is not None: + return self.prev_control_net.doit_ipadapter(model) + else: + return model, [] + + def apply(self, positive, negative, image, mask=None, use_acn=False): + cnet_image_list = [] + prev_cnet_images = [] + + if self.prev_control_net is not None: + positive, negative, prev_cnet_images = self.prev_control_net.apply(positive, negative, image, mask) + + if self.control_image is not None: + cnet_image = self.control_image + elif self.preprocessor is not None: + cnet_image = self.preprocessor.apply(image, mask) + else: + cnet_image = image + + cnet_image_list.extend(prev_cnet_images) + cnet_image_list.append(cnet_image) + + if use_acn: + if "ACN_AdvancedControlNetApply" in nodes.NODE_CLASS_MAPPINGS: + acn = nodes.NODE_CLASS_MAPPINGS['ACN_AdvancedControlNetApply']() + positive, negative, _ = acn.apply_controlnet(positive=positive, negative=negative, control_net=self.control_net, image=cnet_image, + strength=self.strength, start_percent=self.start_percent, end_percent=self.end_percent) + else: + utils.try_install_custom_node('https://github.com/BlenderNeko/ComfyUI_TiledKSampler', + "To use 'ControlNetAdvancedWrapper' for AnimateDiff, 'ComfyUI-Advanced-ControlNet' extension is required.") + raise Exception("'ACN_AdvancedControlNetApply' node isn't installed.") + else: + if self.vae is not None: + apply_controlnet = nodes.ControlNetApplyAdvanced().apply_controlnet + signature = inspect.signature(apply_controlnet) + + if 'vae' in signature.parameters: + positive, negative = nodes.ControlNetApplyAdvanced().apply_controlnet(positive, negative, self.control_net, cnet_image, self.strength, self.start_percent, self.end_percent, vae=self.vae) + else: + print(f"[Impact Pack] ERROR: The ComfyUI version is outdated. VAE cannot be used in ApplyControlNet.") + raise Exception("[Impact Pack] ERROR: The ComfyUI version is outdated. VAE cannot be used in ApplyControlNet.") + else: + positive, negative = nodes.ControlNetApplyAdvanced().apply_controlnet(positive, negative, self.control_net, cnet_image, self.strength, self.start_percent, self.end_percent) + + return positive, negative, cnet_image_list + + +# REQUIREMENTS: BlenderNeko/ComfyUI_TiledKSampler +class TiledKSamplerWrapper: + params = None + + def __init__(self, model, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise, + tile_width, tile_height, tiling_strategy): + self.params = model, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise, tile_width, tile_height, tiling_strategy + + def sample(self, latent_image, hook=None): + if "BNK_TiledKSampler" in nodes.NODE_CLASS_MAPPINGS: + TiledKSampler = nodes.NODE_CLASS_MAPPINGS['BNK_TiledKSampler'] + else: + utils.try_install_custom_node('https://github.com/BlenderNeko/ComfyUI_TiledKSampler', + "To use 'TiledKSamplerProvider', 'Tiled sampling for ComfyUI' extension is required.") + raise Exception("'BNK_TiledKSampler' node isn't installed.") + + model, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise, tile_width, tile_height, tiling_strategy = self.params + + if hook is not None: + model, seed, steps, cfg, sampler_name, scheduler, positive, negative, upscaled_latent, denoise = \ + hook.pre_ksample(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, + denoise) + + return TiledKSampler().sample(model, seed, tile_width, tile_height, tiling_strategy, steps, cfg, sampler_name, + scheduler, positive, negative, latent_image, denoise)[0] + + +class PixelTiledKSampleUpscaler: + def __init__(self, scale_method, model, vae, seed, steps, cfg, sampler_name, scheduler, positive, negative, + denoise, + tile_width, tile_height, tiling_strategy, + upscale_model_opt=None, hook_opt=None, tile_cnet_opt=None, tile_size=512, tile_cnet_strength=1.0): + self.params = scale_method, model, vae, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise + self.vae = vae + self.tile_params = tile_width, tile_height, tiling_strategy + self.upscale_model = upscale_model_opt + self.hook = hook_opt + self.tile_cnet = tile_cnet_opt + self.tile_size = tile_size + self.is_tiled = True + self.tile_cnet_strength = tile_cnet_strength + + def tiled_ksample(self, latent, images): + if "BNK_TiledKSampler" in nodes.NODE_CLASS_MAPPINGS: + TiledKSampler = nodes.NODE_CLASS_MAPPINGS['BNK_TiledKSampler'] + else: + utils.try_install_custom_node('https://github.com/BlenderNeko/ComfyUI_TiledKSampler', + "To use 'PixelTiledKSampleUpscalerProvider', 'Tiled sampling for ComfyUI' extension is required.") + raise RuntimeError("'BNK_TiledKSampler' node isn't installed.") + + scale_method, model, vae, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise = self.params + tile_width, tile_height, tiling_strategy = self.tile_params + + if self.tile_cnet is not None: + image_batch, image_w, image_h, _ = images.shape + if image_batch > 1: + warnings.warn('Multiple latents in batch, Tile ControlNet being ignored') + else: + if 'TilePreprocessor' not in nodes.NODE_CLASS_MAPPINGS: + raise RuntimeError("'TilePreprocessor' node (from comfyui_controlnet_aux) isn't installed.") + preprocessor = nodes.NODE_CLASS_MAPPINGS['TilePreprocessor']() + # might add capacity to set pyrUp_iters later, not needed for now though + preprocessed = preprocessor.execute(images, pyrUp_iters=3, resolution=min(image_w, image_h))[0] + apply_cnet = getattr(nodes.ControlNetApply(), nodes.ControlNetApply.FUNCTION) + positive = apply_cnet(positive, self.tile_cnet, preprocessed, strength=self.tile_cnet_strength)[0] + + return TiledKSampler().sample(model, seed, tile_width, tile_height, tiling_strategy, steps, cfg, sampler_name, + scheduler, positive, negative, latent, denoise)[0] + + def upscale(self, step_info, samples, upscale_factor, save_temp_prefix=None): + scale_method, model, vae, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise = self.params + + if self.hook is not None: + self.hook.set_steps(step_info) + + if self.upscale_model is None: + upscaled_latent, upscaled_images = \ + latent_upscale_on_pixel_space2(samples, scale_method, upscale_factor, vae, + use_tile=True, save_temp_prefix=save_temp_prefix, + hook=self.hook, tile_size=self.tile_size) + else: + upscaled_latent, upscaled_images = \ + latent_upscale_on_pixel_space_with_model2(samples, scale_method, self.upscale_model, + upscale_factor, vae, use_tile=True, + save_temp_prefix=save_temp_prefix, + hook=self.hook, tile_size=self.tile_size) + + refined_latent = self.tiled_ksample(upscaled_latent, upscaled_images) + + return refined_latent + + def upscale_shape(self, step_info, samples, w, h, save_temp_prefix=None): + scale_method, model, vae, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise = self.params + + if self.hook is not None: + self.hook.set_steps(step_info) + + if self.upscale_model is None: + upscaled_latent, upscaled_images = \ + latent_upscale_on_pixel_space_shape2(samples, scale_method, w, h, vae, + use_tile=True, save_temp_prefix=save_temp_prefix, + hook=self.hook, tile_size=self.tile_size) + else: + upscaled_latent, upscaled_images = \ + latent_upscale_on_pixel_space_with_model_shape2(samples, scale_method, + self.upscale_model, w, h, vae, + use_tile=True, + save_temp_prefix=save_temp_prefix, + hook=self.hook, + tile_size=self.tile_size) + + refined_latent = self.tiled_ksample(upscaled_latent, upscaled_images) + + return refined_latent + + +# REQUIREMENTS: biegert/ComfyUI-CLIPSeg +class BBoxDetectorBasedOnCLIPSeg: + prompt = None + blur = None + threshold = None + dilation_factor = None + aux = None + + def __init__(self, prompt, blur, threshold, dilation_factor): + self.prompt = prompt + self.blur = blur + self.threshold = threshold + self.dilation_factor = dilation_factor + + def detect(self, image, bbox_threshold, bbox_dilation, bbox_crop_factor, drop_size=1, detailer_hook=None): + mask = self.detect_combined(image, bbox_threshold, bbox_dilation) + + mask = make_2d_mask(mask) + + segs = mask_to_segs(mask, False, bbox_crop_factor, True, drop_size, detailer_hook=detailer_hook) + + if detailer_hook is not None and hasattr(detailer_hook, "post_detection"): + segs = detailer_hook.post_detection(segs) + + return segs + + def detect_combined(self, image, bbox_threshold, bbox_dilation): + if "CLIPSeg" in nodes.NODE_CLASS_MAPPINGS: + CLIPSeg = nodes.NODE_CLASS_MAPPINGS['CLIPSeg'] + else: + utils.try_install_custom_node('https://github.com/biegert/ComfyUI-CLIPSeg/raw/main/custom_nodes/clipseg.py', + "To use 'CLIPSegDetectorProvider', 'CLIPSeg' extension is required.") + raise Exception("'CLIPSeg' node isn't installed.") + + if self.threshold is None: + threshold = bbox_threshold + else: + threshold = self.threshold + + if self.dilation_factor is None: + dilation_factor = bbox_dilation + else: + dilation_factor = self.dilation_factor + + prompt = self.aux if self.prompt == '' and self.aux is not None else self.prompt + + mask, _, _ = CLIPSeg().segment_image(image, prompt, self.blur, threshold, dilation_factor) + mask = to_binary_mask(mask) + return mask + + def setAux(self, x): + self.aux = x + + +def update_node_status(node, text, progress=None): + if PromptServer.instance.client_id is None: + return + + PromptServer.instance.send_sync("impact/update_status", { + "node": node, + "progress": progress, + "text": text + }, PromptServer.instance.client_id) + + +def random_mask_raw(mask, bbox, factor): + x1, y1, x2, y2 = bbox + w = x2 - x1 + h = y2 - y1 + + factor = max(6, int(min(w, h) * factor / 4)) + + def draw_random_circle(center, radius): + i, j = center + for x in range(int(i - radius), int(i + radius)): + for y in range(int(j - radius), int(j + radius)): + if np.linalg.norm(np.array([x, y]) - np.array([i, j])) <= radius: + mask[x, y] = 1 + + def draw_irregular_line(start, end, pivot, is_vertical): + i = start + while i < end: + base_radius = np.random.randint(5, factor) + radius = int(base_radius) + + if is_vertical: + draw_random_circle((i, pivot), radius) + else: + draw_random_circle((pivot, i), radius) + + i += radius + + def draw_irregular_line_parallel(start, end, pivot, is_vertical): + with ThreadPoolExecutor(max_workers=16) as executor: + futures = [] + step = (end - start) // 16 + for i in range(start, end, step): + future = executor.submit(draw_irregular_line, i, min(i + step, end), pivot, is_vertical) + futures.append(future) + + for future in futures: + future.result() + + draw_irregular_line_parallel(y1 + factor, y2 - factor, x1 + factor, True) + draw_irregular_line_parallel(y1 + factor, y2 - factor, x2 - factor, True) + draw_irregular_line_parallel(x1 + factor, x2 - factor, y1 + factor, False) + draw_irregular_line_parallel(x1 + factor, x2 - factor, y2 - factor, False) + + mask[y1 + factor:y2 - factor, x1 + factor:x2 - factor] = 1.0 + + +def random_mask(mask, bbox, factor, size=128): + small_mask = np.zeros((size, size)).astype(np.float32) + random_mask_raw(small_mask, (0, 0, size, size), factor) + + x1, y1, x2, y2 = bbox + small_mask = torch.tensor(small_mask).unsqueeze(0).unsqueeze(0) + bbox_mask = torch.nn.functional.interpolate(small_mask, size=(y2 - y1, x2 - x1), mode='bilinear', align_corners=False) + bbox_mask = bbox_mask.squeeze(0).squeeze(0) + mask[y1:y2, x1:x2] = bbox_mask + + +def adaptive_mask_paste(dest_mask, src_mask, bbox): + x1, y1, x2, y2 = bbox + small_mask = torch.tensor(src_mask).unsqueeze(0).unsqueeze(0) + bbox_mask = torch.nn.functional.interpolate(small_mask, size=(y2 - y1, x2 - x1), mode='bilinear', align_corners=False) + bbox_mask = bbox_mask.squeeze(0).squeeze(0) + dest_mask[y1:y2, x1:x2] = bbox_mask + + +def crop_condition_mask(mask, image, crop_region): + cond_scale = (mask.shape[1] / image.shape[1], mask.shape[2] / image.shape[2]) + mask_region = [round(v * cond_scale[i % 2]) for i, v in enumerate(crop_region)] + return crop_ndarray3(mask, mask_region) + + +class SafeToGPU: + def __init__(self, size): + self.size = size + + def to_device(self, obj, device): + if utils.is_same_device(device, 'cpu'): + obj.to(device) + else: + if utils.is_same_device(obj.device, 'cpu'): # cpu to gpu + model_management.free_memory(self.size * 1.3, device) + if model_management.get_free_memory(device) > self.size * 1.3: + try: + obj.to(device) + except: + print(f"WARN: The model is not moved to the '{device}' due to insufficient memory. [1]") + else: + print(f"WARN: The model is not moved to the '{device}' due to insufficient memory. [2]") + + +from comfy.cli_args import args, LatentPreviewMethod +import folder_paths +from latent_preview import TAESD, TAESDPreviewerImpl, Latent2RGBPreviewer + +try: + import comfy.latent_formats as latent_formats + + + def get_previewer(device, latent_format=latent_formats.SD15(), force=False, method=None): + previewer = None + + if method is None: + method = args.preview_method + + if method != LatentPreviewMethod.NoPreviews or force: + # TODO previewer methods + taesd_decoder_path = None + + if hasattr(latent_format, "taesd_decoder_path"): + taesd_decoder_path = folder_paths.get_full_path("vae_approx", latent_format.taesd_decoder_name) + + if method == LatentPreviewMethod.Auto: + method = LatentPreviewMethod.Latent2RGB + if taesd_decoder_path: + method = LatentPreviewMethod.TAESD + + if method == LatentPreviewMethod.TAESD: + if taesd_decoder_path: + taesd = TAESD(None, taesd_decoder_path, latent_channels=latent_format.latent_channels).to(device) + previewer = TAESDPreviewerImpl(taesd) + else: + print("Warning: TAESD previews enabled, but could not find models/vae_approx/{}".format( + latent_format.taesd_decoder_name)) + + if previewer is None: + previewer = Latent2RGBPreviewer(latent_format.latent_rgb_factors) + return previewer + +except: + print(f"#########################################################################") + print(f"[ERROR] ComfyUI-Impact-Pack: Please update ComfyUI to the latest version.") + print(f"#########################################################################") diff --git a/ComfyUI-Impact-Pack/modules/impact/defs.py b/ComfyUI-Impact-Pack/modules/impact/defs.py new file mode 100644 index 0000000000000000000000000000000000000000..761a4eb107dfa1562ee7a201a6a611fabdd97d7c --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/defs.py @@ -0,0 +1,17 @@ +detection_labels = [ + 'hand', 'face', 'mouth', 'eyes', 'eyebrows', 'pupils', + 'left_eyebrow', 'left_eye', 'left_pupil', 'right_eyebrow', 'right_eye', 'right_pupil', + 'short_sleeved_shirt', 'long_sleeved_shirt', 'short_sleeved_outwear', 'long_sleeved_outwear', + 'vest', 'sling', 'shorts', 'trousers', 'skirt', 'short_sleeved_dress', 'long_sleeved_dress', 'vest_dress', 'sling_dress', + "person", "bicycle", "car", "motorcycle", "airplane", "bus", "train", "truck", "boat", + "traffic light", "fire hydrant", "stop sign", "parking meter", "bench", + "bird", "cat", "dog", "horse", "sheep", "cow", "elephant", "bear", "zebra", "giraffe", + "backpack", "umbrella", "handbag", "tie", "suitcase", "frisbee", "skis", "snowboard", + "sports ball", "kite", "baseball bat", "baseball glove", "skateboard", "surfboard", + "tennis racket", "bottle", "wine glass", "cup", "fork", "knife", "spoon", "bowl", + "banana", "apple", "sandwich", "orange", "broccoli", "carrot", "hot dog", "pizza", + "donut", "cake", "chair", "couch", "potted plant", "bed", "dining table", "toilet", + "tv", "laptop", "mouse", "remote", "keyboard", "cell phone", "microwave", "oven", + "toaster", "sink", "refrigerator", "book", "clock", "vase", "scissors", "teddy bear", + "hair drier", "toothbrush" + ] \ No newline at end of file diff --git a/ComfyUI-Impact-Pack/modules/impact/detectors.py b/ComfyUI-Impact-Pack/modules/impact/detectors.py new file mode 100644 index 0000000000000000000000000000000000000000..432fcdb60910392d9c84b8c32b6ef7d7b57f8a38 --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/detectors.py @@ -0,0 +1,465 @@ +import impact.core as core +from nodes import MAX_RESOLUTION +import impact.segs_nodes as segs_nodes +import impact.utils as utils +import torch +from impact.core import SEG + +SAM_MODEL_TOOLTIP = {"tooltip": "Segment Anything Model for Silhouette Detection.\nBe sure to use the SAM_MODEL loaded through the SAMLoader (Impact) node as input."} +SAM_MODEL_TOOLTIP_OPTIONAL = {"tooltip": "[OPTIONAL]\nSegment Anything Model for Silhouette Detection.\nBe sure to use the SAM_MODEL loaded through the SAMLoader (Impact) node as input.\nGiven this input, it refines the rectangular areas detected by BBOX_DETECTOR into silhouette shapes through SAM.\nsam_model_opt takes priority over segm_detector_opt."} + +MASK_HINT_THRESHOLD_TOOLTIP = "When detection_hint is mask-area, the mask of SEGS is used as a point hint for SAM (Segment Anything).\nIn this case, only the areas of the mask with brightness values equal to or greater than mask_hint_threshold are used as hints." +MASK_HINT_USE_NEGATIVE_TOOLTIP = "When detecting with SAM (Segment Anything), negative hints are applied as follows:\nSmall: When the SEGS is smaller than 10 pixels in size\nOuter: Sampling the image area outside the SEGS region at regular intervals" + +DILATION_TOOLTIP = "Set the value to dilate the result mask. If the value is negative, it erodes the mask." +DETECTION_HINT_TOOLTIP = {"tooltip": "It is recommended to use only center-1.\nWhen refining the mask of SEGS with the SAM (Segment Anything) model, center-1 uses only the rectangular area of SEGS and a single point at the exact center as hints.\nOther options were added during the experimental stage and do not work well."} + +BBOX_EXPANSION_TOOLTIP = "When performing SAM (Segment Anything) detection within the SEGS area, the rectangular area of SEGS is expanded and used as a hint." + +class SAMDetectorCombined: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "sam_model": ("SAM_MODEL", SAM_MODEL_TOOLTIP), + "segs": ("SEGS", {"tooltip": "This is the segment information detected by the detector.\nIt refines the Mask through the SAM (Segment Anything) detector for all areas pointed to by SEGS, and combines all Masks to return as a single Mask."}), + "image": ("IMAGE", {"tooltip": "It is assumed that segs contains only the information about the detected areas, and does not include the image. SAM (Segment Anything) operates by referencing this image."}), + "detection_hint": (["center-1", "horizontal-2", "vertical-2", "rect-4", "diamond-4", "mask-area", + "mask-points", "mask-point-bbox", "none"], DETECTION_HINT_TOOLTIP), + "dilation": ("INT", {"default": 0, "min": -512, "max": 512, "step": 1, "tooltip": DILATION_TOOLTIP}), + "threshold": ("FLOAT", {"default": 0.93, "min": 0.0, "max": 1.0, "step": 0.01, "tooltip": "Set the sensitivity threshold for the mask detected by SAM (Segment Anything). A higher value generates a more specific mask with a narrower range. For example, when pointing to a person's area, it might detect clothes, which is a narrower range, instead of the entire person."}), + "bbox_expansion": ("INT", {"default": 0, "min": 0, "max": 1000, "step": 1, "tooltip": BBOX_EXPANSION_TOOLTIP}), + "mask_hint_threshold": ("FLOAT", {"default": 0.7, "min": 0.0, "max": 1.0, "step": 0.01, "tooltip": MASK_HINT_THRESHOLD_TOOLTIP}), + "mask_hint_use_negative": (["False", "Small", "Outter"], {"tooltip": MASK_HINT_USE_NEGATIVE_TOOLTIP}) + } + } + + RETURN_TYPES = ("MASK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detector" + + def doit(self, sam_model, segs, image, detection_hint, dilation, + threshold, bbox_expansion, mask_hint_threshold, mask_hint_use_negative): + return (core.make_sam_mask(sam_model, segs, image, detection_hint, dilation, + threshold, bbox_expansion, mask_hint_threshold, mask_hint_use_negative), ) + + +class SAMDetectorSegmented: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "sam_model": ("SAM_MODEL", SAM_MODEL_TOOLTIP), + "segs": ("SEGS", {"tooltip": "This is the segment information detected by the detector.\nFor the SEGS region, the masks detected by SAM (Segment Anything) are created as a unified mask and a batch of individual masks."}), + "image": ("IMAGE", {"tooltip": "It is assumed that segs contains only the information about the detected areas, and does not include the image. SAM (Segment Anything) operates by referencing this image."}), + "detection_hint": (["center-1", "horizontal-2", "vertical-2", "rect-4", "diamond-4", "mask-area", + "mask-points", "mask-point-bbox", "none"], DETECTION_HINT_TOOLTIP), + "dilation": ("INT", {"default": 0, "min": -512, "max": 512, "step": 1, "tooltip": DILATION_TOOLTIP}), + "threshold": ("FLOAT", {"default": 0.93, "min": 0.0, "max": 1.0, "step": 0.01}), + "bbox_expansion": ("INT", {"default": 0, "min": 0, "max": 1000, "step": 1, "tooltip": BBOX_EXPANSION_TOOLTIP}), + "mask_hint_threshold": ("FLOAT", {"default": 0.7, "min": 0.0, "max": 1.0, "step": 0.01, "tooltip": MASK_HINT_THRESHOLD_TOOLTIP}), + "mask_hint_use_negative": (["False", "Small", "Outter"], {"tooltip": MASK_HINT_USE_NEGATIVE_TOOLTIP}) + } + } + + RETURN_TYPES = ("MASK", "MASK") + RETURN_NAMES = ("combined_mask", "batch_masks") + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detector" + + def doit(self, sam_model, segs, image, detection_hint, dilation, + threshold, bbox_expansion, mask_hint_threshold, mask_hint_use_negative): + combined_mask, batch_masks = core.make_sam_mask_segmented(sam_model, segs, image, detection_hint, dilation, + threshold, bbox_expansion, mask_hint_threshold, + mask_hint_use_negative) + return (combined_mask, batch_masks, ) + + +class BboxDetectorForEach: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "bbox_detector": ("BBOX_DETECTOR", ), + "image": ("IMAGE", ), + "threshold": ("FLOAT", {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01}), + "dilation": ("INT", {"default": 10, "min": -512, "max": 512, "step": 1}), + "crop_factor": ("FLOAT", {"default": 3.0, "min": 1.0, "max": 100, "step": 0.1}), + "drop_size": ("INT", {"min": 1, "max": MAX_RESOLUTION, "step": 1, "default": 10}), + "labels": ("STRING", {"multiline": True, "default": "all", "placeholder": "List the types of segments to be allowed, separated by commas"}), + }, + "optional": {"detailer_hook": ("DETAILER_HOOK",), } + } + + RETURN_TYPES = ("SEGS", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detector" + + def doit(self, bbox_detector, image, threshold, dilation, crop_factor, drop_size, labels=None, detailer_hook=None): + if len(image) > 1: + raise Exception('[Impact Pack] ERROR: BboxDetectorForEach does not allow image batches.\nPlease refer to https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/batching-detailer.md for more information.') + + segs = bbox_detector.detect(image, threshold, dilation, crop_factor, drop_size, detailer_hook) + + if labels is not None and labels != '': + labels = labels.split(',') + if len(labels) > 0: + segs, _ = segs_nodes.SEGSLabelFilter.filter(segs, labels) + + return (segs, ) + + +class SegmDetectorForEach: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segm_detector": ("SEGM_DETECTOR", ), + "image": ("IMAGE", ), + "threshold": ("FLOAT", {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01}), + "dilation": ("INT", {"default": 10, "min": -512, "max": 512, "step": 1}), + "crop_factor": ("FLOAT", {"default": 3.0, "min": 1.0, "max": 100, "step": 0.1}), + "drop_size": ("INT", {"min": 1, "max": MAX_RESOLUTION, "step": 1, "default": 10}), + "labels": ("STRING", {"multiline": True, "default": "all", "placeholder": "List the types of segments to be allowed, separated by commas"}), + }, + "optional": {"detailer_hook": ("DETAILER_HOOK",), } + } + + RETURN_TYPES = ("SEGS", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detector" + + def doit(self, segm_detector, image, threshold, dilation, crop_factor, drop_size, labels=None, detailer_hook=None): + if len(image) > 1: + raise Exception('[Impact Pack] ERROR: SegmDetectorForEach does not allow image batches.\nPlease refer to https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/batching-detailer.md for more information.') + + segs = segm_detector.detect(image, threshold, dilation, crop_factor, drop_size, detailer_hook) + + if labels is not None and labels != '': + labels = labels.split(',') + if len(labels) > 0: + segs, _ = segs_nodes.SEGSLabelFilter.filter(segs, labels) + + return (segs, ) + + +class SegmDetectorCombined: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segm_detector": ("SEGM_DETECTOR", ), + "image": ("IMAGE", ), + "threshold": ("FLOAT", {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01}), + "dilation": ("INT", {"default": 0, "min": -512, "max": 512, "step": 1}), + } + } + + RETURN_TYPES = ("MASK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detector" + + def doit(self, segm_detector, image, threshold, dilation): + mask = segm_detector.detect_combined(image, threshold, dilation) + + if mask is None: + mask = torch.zeros((image.shape[2], image.shape[1]), dtype=torch.float32, device="cpu") + + return (mask.unsqueeze(0),) + + +class BboxDetectorCombined(SegmDetectorCombined): + @classmethod + def INPUT_TYPES(s): + return {"required": { + "bbox_detector": ("BBOX_DETECTOR", ), + "image": ("IMAGE", ), + "threshold": ("FLOAT", {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01}), + "dilation": ("INT", {"default": 4, "min": -512, "max": 512, "step": 1}), + } + } + + def doit(self, bbox_detector, image, threshold, dilation): + mask = bbox_detector.detect_combined(image, threshold, dilation) + + if mask is None: + mask = torch.zeros((image.shape[2], image.shape[1]), dtype=torch.float32, device="cpu") + + return (mask.unsqueeze(0),) + + +class SimpleDetectorForEach: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "bbox_detector": ("BBOX_DETECTOR", ), + "image": ("IMAGE", ), + + "bbox_threshold": ("FLOAT", {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01}), + "bbox_dilation": ("INT", {"default": 0, "min": -512, "max": 512, "step": 1}), + + "crop_factor": ("FLOAT", {"default": 3.0, "min": 1.0, "max": 100, "step": 0.1}), + "drop_size": ("INT", {"min": 1, "max": MAX_RESOLUTION, "step": 1, "default": 10}), + + "sub_threshold": ("FLOAT", {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01}), + "sub_dilation": ("INT", {"default": 0, "min": -512, "max": 512, "step": 1}), + "sub_bbox_expansion": ("INT", {"default": 0, "min": 0, "max": 1000, "step": 1}), + + "sam_mask_hint_threshold": ("FLOAT", {"default": 0.7, "min": 0.0, "max": 1.0, "step": 0.01}), + }, + "optional": { + "post_dilation": ("INT", {"default": 0, "min": -512, "max": 512, "step": 1}), + "sam_model_opt": ("SAM_MODEL", SAM_MODEL_TOOLTIP_OPTIONAL), + "segm_detector_opt": ("SEGM_DETECTOR", ), + } + } + + RETURN_TYPES = ("SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detector" + + @staticmethod + def detect(bbox_detector, image, bbox_threshold, bbox_dilation, crop_factor, drop_size, + sub_threshold, sub_dilation, sub_bbox_expansion, + sam_mask_hint_threshold, post_dilation=0, sam_model_opt=None, segm_detector_opt=None, + detailer_hook=None): + if len(image) > 1: + raise Exception('[Impact Pack] ERROR: SimpleDetectorForEach does not allow image batches.\nPlease refer to https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/batching-detailer.md for more information.') + + if segm_detector_opt is not None and hasattr(segm_detector_opt, 'bbox_detector') and segm_detector_opt.bbox_detector == bbox_detector: + # Better segm support for YOLO-World detector + segs = segm_detector_opt.detect(image, sub_threshold, sub_dilation, crop_factor, drop_size, detailer_hook=detailer_hook) + else: + segs = bbox_detector.detect(image, bbox_threshold, bbox_dilation, crop_factor, drop_size, detailer_hook=detailer_hook) + + if sam_model_opt is not None: + mask = core.make_sam_mask(sam_model_opt, segs, image, "center-1", sub_dilation, + sub_threshold, sub_bbox_expansion, sam_mask_hint_threshold, False) + segs = core.segs_bitwise_and_mask(segs, mask) + elif segm_detector_opt is not None: + segm_segs = segm_detector_opt.detect(image, sub_threshold, sub_dilation, crop_factor, drop_size, detailer_hook=detailer_hook) + mask = core.segs_to_combined_mask(segm_segs) + segs = core.segs_bitwise_and_mask(segs, mask) + + segs = core.dilate_segs(segs, post_dilation) + + return (segs,) + + def doit(self, bbox_detector, image, bbox_threshold, bbox_dilation, crop_factor, drop_size, + sub_threshold, sub_dilation, sub_bbox_expansion, + sam_mask_hint_threshold, post_dilation=0, sam_model_opt=None, segm_detector_opt=None): + + return SimpleDetectorForEach.detect(bbox_detector, image, bbox_threshold, bbox_dilation, crop_factor, drop_size, + sub_threshold, sub_dilation, sub_bbox_expansion, + sam_mask_hint_threshold, post_dilation=post_dilation, + sam_model_opt=sam_model_opt, segm_detector_opt=segm_detector_opt) + + +class SimpleDetectorForEachPipe: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "detailer_pipe": ("DETAILER_PIPE", ), + "image": ("IMAGE", ), + + "bbox_threshold": ("FLOAT", {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01}), + "bbox_dilation": ("INT", {"default": 0, "min": -512, "max": 512, "step": 1}), + + "crop_factor": ("FLOAT", {"default": 3.0, "min": 1.0, "max": 100, "step": 0.1}), + "drop_size": ("INT", {"min": 1, "max": MAX_RESOLUTION, "step": 1, "default": 10}), + + "sub_threshold": ("FLOAT", {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01}), + "sub_dilation": ("INT", {"default": 0, "min": -512, "max": 512, "step": 1}), + "sub_bbox_expansion": ("INT", {"default": 0, "min": 0, "max": 1000, "step": 1}), + + "sam_mask_hint_threshold": ("FLOAT", {"default": 0.7, "min": 0.0, "max": 1.0, "step": 0.01}), + }, + "optional": { + "post_dilation": ("INT", {"default": 0, "min": -512, "max": 512, "step": 1}), + } + } + + RETURN_TYPES = ("SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detector" + + def doit(self, detailer_pipe, image, bbox_threshold, bbox_dilation, crop_factor, drop_size, + sub_threshold, sub_dilation, sub_bbox_expansion, sam_mask_hint_threshold, post_dilation=0): + + if len(image) > 1: + raise Exception('[Impact Pack] ERROR: SimpleDetectorForEach does not allow image batches.\nPlease refer to https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/batching-detailer.md for more information.') + + model, clip, vae, positive, negative, wildcard, bbox_detector, segm_detector_opt, sam_model_opt, detailer_hook, refiner_model, refiner_clip, refiner_positive, refiner_negative = detailer_pipe + + return SimpleDetectorForEach.detect(bbox_detector, image, bbox_threshold, bbox_dilation, crop_factor, drop_size, + sub_threshold, sub_dilation, sub_bbox_expansion, + sam_mask_hint_threshold, post_dilation=post_dilation, sam_model_opt=sam_model_opt, segm_detector_opt=segm_detector_opt, + detailer_hook=detailer_hook) + + +class SimpleDetectorForAnimateDiff: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "bbox_detector": ("BBOX_DETECTOR", ), + "image_frames": ("IMAGE", ), + + "bbox_threshold": ("FLOAT", {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01}), + "bbox_dilation": ("INT", {"default": 0, "min": -255, "max": 255, "step": 1}), + + "crop_factor": ("FLOAT", {"default": 3.0, "min": 1.0, "max": 100, "step": 0.1}), + "drop_size": ("INT", {"min": 1, "max": MAX_RESOLUTION, "step": 1, "default": 10}), + + "sub_threshold": ("FLOAT", {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01}), + "sub_dilation": ("INT", {"default": 0, "min": -255, "max": 255, "step": 1}), + "sub_bbox_expansion": ("INT", {"default": 0, "min": 0, "max": 1000, "step": 1}), + + "sam_mask_hint_threshold": ("FLOAT", {"default": 0.7, "min": 0.0, "max": 1.0, "step": 0.01}), + }, + "optional": { + "masking_mode": (["Pivot SEGS", "Combine neighboring frames", "Don't combine"],), + "segs_pivot": (["Combined mask", "1st frame mask"],), + "sam_model_opt": ("SAM_MODEL", SAM_MODEL_TOOLTIP_OPTIONAL), + "segm_detector_opt": ("SEGM_DETECTOR", ), + } + } + + RETURN_TYPES = ("SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detector" + + @staticmethod + def detect(bbox_detector, image_frames, bbox_threshold, bbox_dilation, crop_factor, drop_size, + sub_threshold, sub_dilation, sub_bbox_expansion, sam_mask_hint_threshold, + masking_mode="Pivot SEGS", segs_pivot="Combined mask", sam_model_opt=None, segm_detector_opt=None): + + h = image_frames.shape[1] + w = image_frames.shape[2] + + # gather segs for all frames + segs_by_frames = [] + for image in image_frames: + image = image.unsqueeze(0) + segs = bbox_detector.detect(image, bbox_threshold, bbox_dilation, crop_factor, drop_size) + + if sam_model_opt is not None: + mask = core.make_sam_mask(sam_model_opt, segs, image, "center-1", sub_dilation, + sub_threshold, sub_bbox_expansion, sam_mask_hint_threshold, False) + segs = core.segs_bitwise_and_mask(segs, mask) + elif segm_detector_opt is not None: + segm_segs = segm_detector_opt.detect(image, sub_threshold, sub_dilation, crop_factor, drop_size) + mask = core.segs_to_combined_mask(segm_segs) + segs = core.segs_bitwise_and_mask(segs, mask) + + segs_by_frames.append(segs) + + def get_masked_frames(): + masks_by_frame = [] + for i, segs in enumerate(segs_by_frames): + masks_in_frame = segs_nodes.SEGSToMaskList().doit(segs)[0] + current_frame_mask = (masks_in_frame[0] * 255).to(torch.uint8) + + for mask in masks_in_frame[1:]: + current_frame_mask |= (mask * 255).to(torch.uint8) + + current_frame_mask = (current_frame_mask/255.0).to(torch.float32) + current_frame_mask = utils.to_binary_mask(current_frame_mask, 0.1)[0] + + masks_by_frame.append(current_frame_mask) + + return masks_by_frame + + def get_empty_mask(): + return torch.zeros((h, w), dtype=torch.float32, device="cpu") + + def get_neighboring_mask_at(i, masks_by_frame): + prv = masks_by_frame[i-1] if i > 1 else get_empty_mask() + cur = masks_by_frame[i] + nxt = masks_by_frame[i-1] if i > 1 else get_empty_mask() + + prv = prv if prv is not None else get_empty_mask() + cur = cur.clone() if cur is not None else get_empty_mask() + nxt = nxt if nxt is not None else get_empty_mask() + + return prv, cur, nxt + + def get_merged_neighboring_mask(masks_by_frame): + if len(masks_by_frame) <= 1: + return masks_by_frame + + result = [] + for i in range(0, len(masks_by_frame)): + prv, cur, nxt = get_neighboring_mask_at(i, masks_by_frame) + cur = (cur * 255).to(torch.uint8) + cur |= (prv * 255).to(torch.uint8) + cur |= (nxt * 255).to(torch.uint8) + cur = (cur / 255.0).to(torch.float32) + cur = utils.to_binary_mask(cur, 0.1)[0] + result.append(cur) + + return result + + def get_whole_merged_mask(): + all_masks = [] + for segs in segs_by_frames: + all_masks += segs_nodes.SEGSToMaskList().doit(segs)[0] + + merged_mask = (all_masks[0] * 255).to(torch.uint8) + for mask in all_masks[1:]: + merged_mask |= (mask * 255).to(torch.uint8) + + merged_mask = (merged_mask / 255.0).to(torch.float32) + merged_mask = utils.to_binary_mask(merged_mask, 0.1)[0] + return merged_mask + + def get_pivot_segs(): + if segs_pivot == "1st frame mask": + return segs_by_frames[0][1] + else: + merged_mask = get_whole_merged_mask() + return segs_nodes.MaskToSEGS.doit(merged_mask, False, crop_factor, False, drop_size, contour_fill=True)[0] + + def get_segs(merged_neighboring=False): + pivot_segs = get_pivot_segs() + + masks_by_frame = get_masked_frames() + if merged_neighboring: + masks_by_frame = get_merged_neighboring_mask(masks_by_frame) + + new_segs = [] + for seg in pivot_segs[1]: + cropped_mask = torch.zeros(seg.cropped_mask.shape, dtype=torch.float32, device="cpu").unsqueeze(0) + pivot_mask = torch.from_numpy(seg.cropped_mask) + x1, y1, x2, y2 = seg.crop_region + for mask in masks_by_frame: + cropped_mask_at_frame = (mask[y1:y2, x1:x2] * pivot_mask).unsqueeze(0) + cropped_mask = torch.cat((cropped_mask, cropped_mask_at_frame), dim=0) + + if len(cropped_mask) > 1: + cropped_mask = cropped_mask[1:] + + new_seg = SEG(seg.cropped_image, cropped_mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, seg.control_net_wrapper) + new_segs.append(new_seg) + + return pivot_segs[0], new_segs + + # create result mask + if masking_mode == "Pivot SEGS": + return (get_pivot_segs(), ) + + elif masking_mode == "Combine neighboring frames": + return (get_segs(merged_neighboring=True), ) + + else: # elif masking_mode == "Don't combine": + return (get_segs(merged_neighboring=False), ) + + def doit(self, bbox_detector, image_frames, bbox_threshold, bbox_dilation, crop_factor, drop_size, + sub_threshold, sub_dilation, sub_bbox_expansion, sam_mask_hint_threshold, + masking_mode="Pivot SEGS", segs_pivot="Combined mask", sam_model_opt=None, segm_detector_opt=None): + + return SimpleDetectorForAnimateDiff.detect(bbox_detector, image_frames, bbox_threshold, bbox_dilation, crop_factor, drop_size, + sub_threshold, sub_dilation, sub_bbox_expansion, sam_mask_hint_threshold, + masking_mode, segs_pivot, sam_model_opt, segm_detector_opt) diff --git a/ComfyUI-Impact-Pack/modules/impact/hf_nodes.py b/ComfyUI-Impact-Pack/modules/impact/hf_nodes.py new file mode 100644 index 0000000000000000000000000000000000000000..b93e040e55225e532faf5afa064b56b87e39e177 --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/hf_nodes.py @@ -0,0 +1,188 @@ +import comfy +import re +from impact.utils import * + +hf_transformer_model_urls = [ + "rizvandwiki/gender-classification-2", + "NTQAI/pedestrian_gender_recognition", + "Leilab/gender_class", + "ProjectPersonal/GenderClassifier", + "crangana/trained-gender", + "cledoux42/GenderNew_v002", + "ivensamdh/genderage2" +] + + +class HF_TransformersClassifierProvider: + @classmethod + def INPUT_TYPES(s): + global hf_transformer_model_urls + return {"required": { + "preset_repo_id": (hf_transformer_model_urls + ['Manual repo id'],), + "manual_repo_id": ("STRING", {"multiline": False}), + "device_mode": (["AUTO", "Prefer GPU", "CPU"],), + }, + } + + RETURN_TYPES = ("TRANSFORMERS_CLASSIFIER",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/HuggingFace" + + def doit(self, preset_repo_id, manual_repo_id, device_mode): + from transformers import pipeline + + if preset_repo_id == 'Manual repo id': + url = manual_repo_id + else: + url = preset_repo_id + + if device_mode != 'CPU': + device = comfy.model_management.get_torch_device() + else: + device = "cpu" + + classifier = pipeline('image-classification', model=url, device=device) + + return (classifier,) + + +preset_classify_expr = [ + '#Female > #Male', + '#Female < #Male', + 'female > 0.5', + 'male > 0.5', + 'Age16to25 > 0.1', + 'Age50to69 > 0.1', +] + +symbolic_label_map = { + '#Female': {'female', 'Female', 'Human Female', 'woman', 'women', 'girl'}, + '#Male': {'male', 'Male', 'Human Male', 'man', 'men', 'boy'} +} + +def is_numeric_string(input_str): + return re.match(r'^-?\d+(\.\d+)?$', input_str) is not None + + +classify_expr_pattern = r'([^><= ]+)\s*(>|<|>=|<=|=)\s*([^><= ]+)' + + +class SEGS_Classify: + @classmethod + def INPUT_TYPES(s): + global preset_classify_expr + return {"required": { + "classifier": ("TRANSFORMERS_CLASSIFIER",), + "segs": ("SEGS",), + "preset_expr": (preset_classify_expr + ['Manual expr'],), + "manual_expr": ("STRING", {"multiline": False}), + }, + "optional": { + "ref_image_opt": ("IMAGE", ), + } + } + + RETURN_TYPES = ("SEGS", "SEGS", "STRING") + RETURN_NAMES = ("filtered_SEGS", "remained_SEGS", "detected_labels") + OUTPUT_IS_LIST = (False, False, True) + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/HuggingFace" + + @staticmethod + def lookup_classified_label_score(score_infos, label): + global symbolic_label_map + + if label.startswith('#'): + if label not in symbolic_label_map: + return None + else: + label = symbolic_label_map[label] + else: + label = {label} + + for x in score_infos: + if x['label'] in label: + return x['score'] + + return None + + def doit(self, classifier, segs, preset_expr, manual_expr, ref_image_opt=None): + if preset_expr == 'Manual expr': + expr_str = manual_expr + else: + expr_str = preset_expr + + match = re.match(classify_expr_pattern, expr_str) + + if match is None: + return (segs[0], []), segs, [] + + a = match.group(1) + op = match.group(2) + b = match.group(3) + + a_is_lab = not is_numeric_string(a) + b_is_lab = not is_numeric_string(b) + + classified = [] + remained_SEGS = [] + provided_labels = set() + + for seg in segs[1]: + cropped_image = None + + if seg.cropped_image is not None: + cropped_image = seg.cropped_image + elif ref_image_opt is not None: + # take from original image + cropped_image = crop_image(ref_image_opt, seg.crop_region) + + if cropped_image is not None: + cropped_image = to_pil(cropped_image) + res = classifier(cropped_image) + classified.append((seg, res)) + + for x in res: + provided_labels.add(x['label']) + else: + remained_SEGS.append(seg) + + filtered_SEGS = [] + for seg, res in classified: + if a_is_lab: + avalue = SEGS_Classify.lookup_classified_label_score(res, a) + else: + avalue = a + + if b_is_lab: + bvalue = SEGS_Classify.lookup_classified_label_score(res, b) + else: + bvalue = b + + if avalue is None or bvalue is None: + remained_SEGS.append(seg) + continue + + avalue = float(avalue) + bvalue = float(bvalue) + + if op == '>': + cond = avalue > bvalue + elif op == '<': + cond = avalue < bvalue + elif op == '>=': + cond = avalue >= bvalue + elif op == '<=': + cond = avalue <= bvalue + else: + cond = avalue == bvalue + + if cond: + filtered_SEGS.append(seg) + else: + remained_SEGS.append(seg) + + return (segs[0], filtered_SEGS), (segs[0], remained_SEGS), list(provided_labels) diff --git a/ComfyUI-Impact-Pack/modules/impact/hook_nodes.py b/ComfyUI-Impact-Pack/modules/impact/hook_nodes.py new file mode 100644 index 0000000000000000000000000000000000000000..c72658fcae32c108a532925c665458359f7e73cc --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/hook_nodes.py @@ -0,0 +1,85 @@ +import sys +from . import hooks +from . import defs + + +class SEGSOrderedFilterDetailerHookProvider: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "target": (["area(=w*h)", "width", "height", "x1", "y1", "x2", "y2"],), + "order": ("BOOLEAN", {"default": True, "label_on": "descending", "label_off": "ascending"}), + "take_start": ("INT", {"default": 0, "min": 0, "max": sys.maxsize, "step": 1}), + "take_count": ("INT", {"default": 1, "min": 0, "max": sys.maxsize, "step": 1}), + }, + } + + RETURN_TYPES = ("DETAILER_HOOK", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, target, order, take_start, take_count): + hook = hooks.SEGSOrderedFilterDetailerHook(target, order, take_start, take_count) + return (hook, ) + + +class SEGSRangeFilterDetailerHookProvider: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "target": (["area(=w*h)", "width", "height", "x1", "y1", "x2", "y2", "length_percent"],), + "mode": ("BOOLEAN", {"default": True, "label_on": "inside", "label_off": "outside"}), + "min_value": ("INT", {"default": 0, "min": 0, "max": sys.maxsize, "step": 1}), + "max_value": ("INT", {"default": 67108864, "min": 0, "max": sys.maxsize, "step": 1}), + }, + } + + RETURN_TYPES = ("DETAILER_HOOK", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, target, mode, min_value, max_value): + hook = hooks.SEGSRangeFilterDetailerHook(target, mode, min_value, max_value) + return (hook, ) + + +class SEGSLabelFilterDetailerHookProvider: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS", ), + "preset": (['all'] + defs.detection_labels,), + "labels": ("STRING", {"multiline": True, "placeholder": "List the types of segments to be allowed, separated by commas"}), + }, + } + + RETURN_TYPES = ("DETAILER_HOOK", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, preset, labels): + hook = hooks.SEGSLabelFilterDetailerHook(labels) + return (hook, ) + + +class PreviewDetailerHookProvider: + @classmethod + def INPUT_TYPES(s): + return { + "required": {"quality": ("INT", {"default": 95, "min": 20, "max": 100})}, + "hidden": {"unique_id": "UNIQUE_ID"}, + } + + RETURN_TYPES = ("DETAILER_HOOK", "UPSCALER_HOOK") + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + NOT_IDEMPOTENT = True + + def doit(self, quality, unique_id): + hook = hooks.PreviewDetailerHook(unique_id, quality) + return hook, hook diff --git a/ComfyUI-Impact-Pack/modules/impact/hooks.py b/ComfyUI-Impact-Pack/modules/impact/hooks.py new file mode 100644 index 0000000000000000000000000000000000000000..393fae500fac9a245b1aec37ba2837d0e55efd1e --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/hooks.py @@ -0,0 +1,518 @@ +import copy +import torch +import nodes +from impact import utils +from . import segs_nodes +from thirdparty import noise_nodes +from server import PromptServer +import asyncio +import folder_paths +import os +from comfy_extras import nodes_custom_sampler +import math + + +class PixelKSampleHook: + cur_step = 0 + total_step = 0 + + def __init__(self): + pass + + def set_steps(self, info): + self.cur_step, self.total_step = info + + def post_decode(self, pixels): + return pixels + + def post_upscale(self, pixels): + return pixels + + def post_encode(self, samples): + return samples + + def pre_decode(self, samples): + return samples + + def pre_ksample(self, model, seed, steps, cfg, sampler_name, scheduler, positive, negative, upscaled_latent, + denoise): + return model, seed, steps, cfg, sampler_name, scheduler, positive, negative, upscaled_latent, denoise + + def post_crop_region(self, w, h, item_bbox, crop_region): + return crop_region + + def touch_scaled_size(self, w, h): + return w, h + + +class PixelKSampleHookCombine(PixelKSampleHook): + hook1 = None + hook2 = None + + def __init__(self, hook1, hook2): + super().__init__() + self.hook1 = hook1 + self.hook2 = hook2 + + def set_steps(self, info): + self.hook1.set_steps(info) + self.hook2.set_steps(info) + + def pre_decode(self, samples): + return self.hook2.pre_decode(self.hook1.pre_decode(samples)) + + def post_decode(self, pixels): + return self.hook2.post_decode(self.hook1.post_decode(pixels)) + + def post_upscale(self, pixels): + return self.hook2.post_upscale(self.hook1.post_upscale(pixels)) + + def post_encode(self, samples): + return self.hook2.post_encode(self.hook1.post_encode(samples)) + + def post_crop_region(self, w, h, item_bbox, crop_region): + crop_region = self.hook1.post_crop_region(w, h, item_bbox, crop_region) + return self.hook2.post_crop_region(w, h, item_bbox, crop_region) + + def touch_scaled_size(self, w, h): + w, h = self.hook1.touch_scaled_size(w, h) + return self.hook2.touch_scaled_size(w, h) + + def pre_ksample(self, model, seed, steps, cfg, sampler_name, scheduler, positive, negative, upscaled_latent, + denoise): + model, seed, steps, cfg, sampler_name, scheduler, positive, negative, upscaled_latent, denoise = \ + self.hook1.pre_ksample(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, + upscaled_latent, denoise) + + return self.hook2.pre_ksample(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, + upscaled_latent, denoise) + + +class DetailerHookCombine(PixelKSampleHookCombine): + def cycle_latent(self, latent): + latent = self.hook1.cycle_latent(latent) + latent = self.hook2.cycle_latent(latent) + return latent + + def post_detection(self, segs): + segs = self.hook1.post_detection(segs) + segs = self.hook2.post_detection(segs) + return segs + + def post_paste(self, image): + image = self.hook1.post_paste(image) + image = self.hook2.post_paste(image) + return image + + def get_custom_noise(self, seed, noise, is_touched): + noise_1st, is_touched = self.hook1.get_custom_noise(seed, noise, is_touched) + noise_2nd, is_touched = self.hook2.get_custom_noise(seed, noise, is_touched) + return noise, is_touched + + +class SimpleCfgScheduleHook(PixelKSampleHook): + target_cfg = 0 + + def __init__(self, target_cfg): + super().__init__() + self.target_cfg = target_cfg + + def pre_ksample(self, model, seed, steps, cfg, sampler_name, scheduler, positive, negative, upscaled_latent, denoise): + if self.total_step > 1: + progress = self.cur_step / (self.total_step - 1) + gap = self.target_cfg - cfg + current_cfg = int(cfg + gap * progress) + else: + current_cfg = self.target_cfg + + return model, seed, steps, current_cfg, sampler_name, scheduler, positive, negative, upscaled_latent, denoise + + +class SimpleDenoiseScheduleHook(PixelKSampleHook): + def __init__(self, target_denoise): + super().__init__() + self.target_denoise = target_denoise + + def pre_ksample(self, model, seed, steps, cfg, sampler_name, scheduler, positive, negative, upscaled_latent, denoise): + if self.total_step > 1: + progress = self.cur_step / (self.total_step - 1) + gap = self.target_denoise - denoise + current_denoise = denoise + gap * progress + else: + current_denoise = self.target_denoise + + return model, seed, steps, cfg, sampler_name, scheduler, positive, negative, upscaled_latent, current_denoise + + +class SimpleStepsScheduleHook(PixelKSampleHook): + def __init__(self, target_steps): + super().__init__() + self.target_steps = target_steps + + def pre_ksample(self, model, seed, steps, cfg, sampler_name, scheduler, positive, negative, upscaled_latent, denoise): + if self.total_step > 1: + progress = self.cur_step / (self.total_step - 1) + gap = self.target_steps - steps + current_steps = int(steps + gap * progress) + else: + current_steps = self.target_steps + + return model, seed, current_steps, cfg, sampler_name, scheduler, positive, negative, upscaled_latent, denoise + + +class DetailerHook(PixelKSampleHook): + def cycle_latent(self, latent): + return latent + + def post_detection(self, segs): + return segs + + def post_paste(self, image): + return image + + def get_custom_noise(self, seed, noise, is_touched): + return noise, is_touched + + +# class CustomNoiseDetailerHookProvider(DetailerHook): +# def __init__(self, noise): +# super().__init__() +# self.noise = noise +# +# def get_custom_noise(self, seed, noise, is_start): +# return self.noise + + +class VariationNoiseDetailerHookProvider(DetailerHook): + def __init__(self, variation_seed, variation_strength): + super().__init__() + self.variation_seed = variation_seed + self.variation_strength = variation_strength + + def get_custom_noise(self, seed, noise, is_touched): + empty_noise = {'samples': torch.zeros(noise.size())} + if not is_touched: + noise = nodes_custom_sampler.Noise_RandomNoise(seed).generate_noise(empty_noise) + noise_2nd = nodes_custom_sampler.Noise_RandomNoise(self.variation_seed).generate_noise(empty_noise) + + mixed_noise = ((1 - self.variation_strength) * noise + self.variation_strength * noise_2nd) + + # NOTE: Since the variance of the Gaussian noise in mixed_noise has changed, it must be corrected through scaling. + scale_factor = math.sqrt((1 - self.variation_strength) ** 2 + self.variation_strength ** 2) + corrected_noise = mixed_noise / scale_factor # Scale the noise to maintain variance of 1 + + return corrected_noise, True + + +class SimpleDetailerDenoiseSchedulerHook(DetailerHook): + def __init__(self, target_denoise): + super().__init__() + self.target_denoise = target_denoise + + def pre_ksample(self, model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent, denoise): + if self.total_step > 1: + progress = self.cur_step / (self.total_step - 1) + gap = self.target_denoise - denoise + current_denoise = denoise + gap * progress + else: + # ignore hook if total cycle <= 1 + current_denoise = denoise + + return model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent, current_denoise + + +class CoreMLHook(DetailerHook): + def __init__(self, mode): + super().__init__() + resolution = mode.split('x') + + self.w = int(resolution[0]) + self.h = int(resolution[1]) + + self.override_bbox_by_segm = False + + def pre_decode(self, samples): + new_samples = copy.deepcopy(samples) + new_samples['samples'] = samples['samples'][0].unsqueeze(0) + return new_samples + + def post_encode(self, samples): + new_samples = copy.deepcopy(samples) + new_samples['samples'] = samples['samples'].repeat(2, 1, 1, 1) + return new_samples + + def post_crop_region(self, w, h, item_bbox, crop_region): + x1, y1, x2, y2 = crop_region + bx1, by1, bx2, by2 = item_bbox + crop_w = x2-x1 + crop_h = y2-y1 + + crop_ratio = crop_w/crop_h + target_ratio = self.w/self.h + if crop_ratio < target_ratio: + # shrink height + top_gap = by1 - y1 + bottom_gap = y2 - by2 + + gap_ratio = top_gap / bottom_gap + + target_height = 1/target_ratio*crop_w + delta_height = crop_h - target_height + + new_y1 = int(y1 + delta_height*gap_ratio) + new_y2 = int(new_y1 + target_height) + crop_region = x1, new_y1, x2, new_y2 + + elif crop_ratio > target_ratio: + # shrink width + left_gap = bx1 - x1 + right_gap = x2 - bx2 + + gap_ratio = left_gap / right_gap + + target_width = target_ratio*crop_h + delta_width = crop_w - target_width + + new_x1 = int(x1 + delta_width*gap_ratio) + new_x2 = int(new_x1 + target_width) + crop_region = new_x1, y1, new_x2, y2 + + return crop_region + + def touch_scaled_size(self, w, h): + return self.w, self.h + + +# REQUIREMENTS: BlenderNeko/ComfyUI Noise +class InjectNoiseHook(PixelKSampleHook): + def __init__(self, source, seed, start_strength, end_strength): + super().__init__() + self.source = source + self.seed = seed + self.start_strength = start_strength + self.end_strength = end_strength + + def post_encode(self, samples): + cur_step = self.cur_step + + size = samples['samples'].shape + seed = cur_step + self.seed + cur_step + + if "BNK_NoisyLatentImage" in nodes.NODE_CLASS_MAPPINGS and "BNK_InjectNoise" in nodes.NODE_CLASS_MAPPINGS: + NoisyLatentImage = nodes.NODE_CLASS_MAPPINGS["BNK_NoisyLatentImage"] + InjectNoise = nodes.NODE_CLASS_MAPPINGS["BNK_InjectNoise"] + else: + utils.try_install_custom_node('https://github.com/BlenderNeko/ComfyUI_Noise', + "To use 'NoiseInjectionHookProvider', 'ComfyUI Noise' extension is required.") + raise Exception("'BNK_NoisyLatentImage', 'BNK_InjectNoise' nodes are not installed.") + + noise = NoisyLatentImage().create_noisy_latents(self.source, seed, size[3] * 8, size[2] * 8, size[0])[0] + + # inj noise + mask = None + if 'noise_mask' in samples: + mask = samples['noise_mask'] + + strength = self.start_strength + (self.end_strength - self.start_strength) * cur_step / self.total_step + samples = InjectNoise().inject_noise(samples, strength, noise, mask)[0] + print(f"[Impact Pack] InjectNoiseHook: strength = {strength}") + + if mask is not None: + samples['noise_mask'] = mask + + return samples + + +class UnsamplerHook(PixelKSampleHook): + def __init__(self, model, steps, start_end_at_step, end_end_at_step, cfg, sampler_name, + scheduler, normalize, positive, negative): + super().__init__() + self.model = model + self.cfg = cfg + self.sampler_name = sampler_name + self.steps = steps + self.start_end_at_step = start_end_at_step + self.end_end_at_step = end_end_at_step + self.scheduler = scheduler + self.normalize = normalize + self.positive = positive + self.negative = negative + + def post_encode(self, samples): + cur_step = self.cur_step + + Unsampler = noise_nodes.Unsampler + + end_at_step = self.start_end_at_step + (self.end_end_at_step - self.start_end_at_step) * cur_step / self.total_step + end_at_step = int(end_at_step) + + print(f"[Impact Pack] UnsamplerHook: end_at_step = {end_at_step}") + + # inj noise + mask = None + if 'noise_mask' in samples: + mask = samples['noise_mask'] + + samples = Unsampler().unsampler(self.model, self.cfg, self.sampler_name, self.steps, end_at_step, + self.scheduler, self.normalize, self.positive, self.negative, samples)[0] + + if mask is not None: + samples['noise_mask'] = mask + + return samples + + +class InjectNoiseHookForDetailer(DetailerHook): + def __init__(self, source, seed, start_strength, end_strength, from_start=False): + super().__init__() + self.source = source + self.seed = seed + self.start_strength = start_strength + self.end_strength = end_strength + self.from_start = from_start + + def inject_noise(self, samples): + cur_step = self.cur_step if self.from_start else self.cur_step - 1 + total_step = self.total_step if self.from_start else self.total_step - 1 + + size = samples['samples'].shape + seed = cur_step + self.seed + cur_step + + if "BNK_NoisyLatentImage" in nodes.NODE_CLASS_MAPPINGS and "BNK_InjectNoise" in nodes.NODE_CLASS_MAPPINGS: + NoisyLatentImage = nodes.NODE_CLASS_MAPPINGS["BNK_NoisyLatentImage"] + InjectNoise = nodes.NODE_CLASS_MAPPINGS["BNK_InjectNoise"] + else: + utils.try_install_custom_node('https://github.com/BlenderNeko/ComfyUI_Noise', + "To use 'NoiseInjectionDetailerHookProvider', 'ComfyUI Noise' extension is required.") + raise Exception("'BNK_NoisyLatentImage', 'BNK_InjectNoise' nodes are not installed.") + + noise = NoisyLatentImage().create_noisy_latents(self.source, seed, size[3] * 8, size[2] * 8, size[0])[0] + + # inj noise + mask = None + if 'noise_mask' in samples: + mask = samples['noise_mask'] + + strength = self.start_strength + (self.end_strength - self.start_strength) * cur_step / total_step + samples = InjectNoise().inject_noise(samples, strength, noise, mask)[0] + + if mask is not None: + samples['noise_mask'] = mask + + return samples + + def cycle_latent(self, latent): + if self.cur_step == 0 and not self.from_start: + return latent + else: + return self.inject_noise(latent) + + +class UnsamplerDetailerHook(DetailerHook): + def __init__(self, model, steps, start_end_at_step, end_end_at_step, cfg, sampler_name, + scheduler, normalize, positive, negative, from_start=False): + super().__init__() + self.model = model + self.cfg = cfg + self.sampler_name = sampler_name + self.steps = steps + self.start_end_at_step = start_end_at_step + self.end_end_at_step = end_end_at_step + self.scheduler = scheduler + self.normalize = normalize + self.positive = positive + self.negative = negative + self.from_start = from_start + + def unsample(self, samples): + cur_step = self.cur_step if self.from_start else self.cur_step - 1 + total_step = self.total_step if self.from_start else self.total_step - 1 + + Unsampler = noise_nodes.Unsampler + + end_at_step = self.start_end_at_step + (self.end_end_at_step - self.start_end_at_step) * cur_step / total_step + end_at_step = int(end_at_step) + + # inj noise + mask = None + if 'noise_mask' in samples: + mask = samples['noise_mask'] + + samples = Unsampler().unsampler(self.model, self.cfg, self.sampler_name, self.steps, end_at_step, + self.scheduler, self.normalize, self.positive, self.negative, samples)[0] + + if mask is not None: + samples['noise_mask'] = mask + + return samples + + def cycle_latent(self, latent): + if self.cur_step == 0 and not self.from_start: + return latent + else: + return self.unsample(latent) + + +class SEGSOrderedFilterDetailerHook(DetailerHook): + def __init__(self, target, order, take_start, take_count): + super().__init__() + self.target = target + self.order = order + self.take_start = take_start + self.take_count = take_count + + def post_detection(self, segs): + return segs_nodes.SEGSOrderedFilter().doit(segs, self.target, self.order, self.take_start, self.take_count)[0] + + +class SEGSRangeFilterDetailerHook(DetailerHook): + def __init__(self, target, mode, min_value, max_value): + super().__init__() + self.target = target + self.mode = mode + self.min_value = min_value + self.max_value = max_value + + def post_detection(self, segs): + return segs_nodes.SEGSRangeFilter().doit(segs, self.target, self.mode, self.min_value, self.max_value)[0] + + +class SEGSLabelFilterDetailerHook(DetailerHook): + def __init__(self, labels): + super().__init__() + self.labels = labels + + def post_detection(self, segs): + return segs_nodes.SEGSLabelFilter().doit(segs, "", self.labels)[0] + + +class PreviewDetailerHook(DetailerHook): + def __init__(self, node_id, quality): + super().__init__() + self.node_id = node_id + self.quality = quality + + async def send(self, image): + if len(image) > 0: + image = image[0].unsqueeze(0) + img = utils.tensor2pil(image) + + temp_path = os.path.join(folder_paths.get_temp_directory(), 'pvhook') + + if not os.path.exists(temp_path): + os.makedirs(temp_path) + + fullpath = os.path.join(temp_path, f"{self.node_id}.webp") + img.save(fullpath, quality=self.quality) + + item = { + "filename": f"{self.node_id}.webp", + "subfolder": 'pvhook', + "type": 'temp' + } + + PromptServer.instance.send_sync("impact-preview", {'node_id': self.node_id, 'item': item}) + + def post_paste(self, image): + asyncio.run(self.send(image)) + return image diff --git a/ComfyUI-Impact-Pack/modules/impact/impact_pack.py b/ComfyUI-Impact-Pack/modules/impact/impact_pack.py new file mode 100644 index 0000000000000000000000000000000000000000..6179ee2f58cd68f190eb75fd6d9efeef079ca257 --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/impact_pack.py @@ -0,0 +1,2231 @@ +import os +import sys + +import comfy.samplers +import comfy.sd +import warnings +from segment_anything import sam_model_registry +from io import BytesIO +import piexif +import zipfile +import re + +import impact.wildcards + +from impact.utils import * +import impact.core as core +from impact.core import SEG +from impact.config import latent_letter_path +from nodes import MAX_RESOLUTION +from PIL import Image, ImageOps +import numpy as np +import hashlib +import json +import safetensors.torch +from PIL.PngImagePlugin import PngInfo +import comfy.model_management +import base64 +import impact.wildcards as wildcards +from . import hooks +from . import utils + + +try: + from comfy_extras import nodes_differential_diffusion +except Exception: + print(f"\n#############################################\n[Impact Pack] ComfyUI is an outdated version.\n#############################################\n") + raise Exception("[Impact Pack] ComfyUI is an outdated version.") + + +warnings.filterwarnings('ignore', category=UserWarning, message='TypedStorage is deprecated') + +model_path = folder_paths.models_dir + + +# folder_paths.supported_pt_extensions +add_folder_path_and_extensions("mmdets_bbox", [os.path.join(model_path, "mmdets", "bbox")], folder_paths.supported_pt_extensions) +add_folder_path_and_extensions("mmdets_segm", [os.path.join(model_path, "mmdets", "segm")], folder_paths.supported_pt_extensions) +add_folder_path_and_extensions("mmdets", [os.path.join(model_path, "mmdets")], folder_paths.supported_pt_extensions) +add_folder_path_and_extensions("sams", [os.path.join(model_path, "sams")], folder_paths.supported_pt_extensions) +add_folder_path_and_extensions("onnx", [os.path.join(model_path, "onnx")], {'.onnx'}) + + +# Nodes +class ONNXDetectorProvider: + @classmethod + def INPUT_TYPES(s): + return {"required": {"model_name": (folder_paths.get_filename_list("onnx"), )}} + + RETURN_TYPES = ("BBOX_DETECTOR", ) + FUNCTION = "load_onnx" + + CATEGORY = "ImpactPack" + + def load_onnx(self, model_name): + model = folder_paths.get_full_path("onnx", model_name) + return (core.ONNXDetector(model), ) + + +class CLIPSegDetectorProvider: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "text": ("STRING", {"multiline": False, "tooltip": "Enter the targets to be detected, separated by commas"}), + "blur": ("FLOAT", {"min": 0, "max": 15, "step": 0.1, "default": 7, "tooltip": "Blurs the detected mask"}), + "threshold": ("FLOAT", {"min": 0, "max": 1, "step": 0.05, "default": 0.4, "tooltip": "Detects only areas that are certain above the threshold."}), + "dilation_factor": ("INT", {"min": 0, "max": 10, "step": 1, "default": 4, "tooltip": "Dilates the detected mask."}), + } + } + + RETURN_TYPES = ("BBOX_DETECTOR", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + DESCRIPTION = "Provides a detection function using CLIPSeg, which generates masks based on text prompts.\nTo use this node, the CLIPSeg custom node must be installed." + + def doit(self, text, blur, threshold, dilation_factor): + if "CLIPSeg" in nodes.NODE_CLASS_MAPPINGS: + return (core.BBoxDetectorBasedOnCLIPSeg(text, blur, threshold, dilation_factor), ) + else: + print("[ERROR] CLIPSegToBboxDetector: CLIPSeg custom node isn't installed. You must install biegert/ComfyUI-CLIPSeg extension to use this node.") + + +class SAMLoader: + @classmethod + def INPUT_TYPES(cls): + models = [x for x in folder_paths.get_filename_list("sams") if 'hq' not in x] + return { + "required": { + "model_name": (models + ['ESAM'], {"tooltip": "The detection accuracy varies depending on the SAM model. ESAM can only be used if ComfyUI-YoloWorld-EfficientSAM is installed."}), + "device_mode": (["AUTO", "Prefer GPU", "CPU"], {"tooltip": "AUTO: Only applicable when a GPU is available. It temporarily loads the SAM_MODEL into VRAM only when the detection function is used.\n" + "Prefer GPU: Tries to keep the SAM_MODEL on the GPU whenever possible. This can be used when there is sufficient VRAM available.\n" + "CPU: Always loads only on the CPU."}), + } + } + + RETURN_TYPES = ("SAM_MODEL", ) + FUNCTION = "load_model" + + CATEGORY = "ImpactPack" + + DESCRIPTION = "Load the SAM (Segment Anything) model. This can be used in places that utilize SAM detection functionality, such as SAMDetector or SimpleDetector.\nThe SAM detection functionality in Impact Pack must use the SAM_MODEL loaded through this node." + + def load_model(self, model_name, device_mode="auto"): + if model_name == 'ESAM': + if 'ESAM_ModelLoader_Zho' not in nodes.NODE_CLASS_MAPPINGS: + try_install_custom_node('https://github.com/ZHO-ZHO-ZHO/ComfyUI-YoloWorld-EfficientSAM', + "To use 'ESAM' model, 'ComfyUI-YoloWorld-EfficientSAM' extension is required.") + raise Exception("'ComfyUI-YoloWorld-EfficientSAM' node isn't installed.") + + esam_loader = nodes.NODE_CLASS_MAPPINGS['ESAM_ModelLoader_Zho']() + + if device_mode == 'CPU': + esam = esam_loader.load_esam_model('CPU')[0] + else: + device_mode = 'CUDA' + esam = esam_loader.load_esam_model('CUDA')[0] + + sam_obj = core.ESAMWrapper(esam, device_mode) + esam.sam_wrapper = sam_obj + + print(f"Loads EfficientSAM model: (device:{device_mode})") + return (esam, ) + + modelname = folder_paths.get_full_path("sams", model_name) + + if 'vit_h' in model_name: + model_kind = 'vit_h' + elif 'vit_l' in model_name: + model_kind = 'vit_l' + else: + model_kind = 'vit_b' + + sam = sam_model_registry[model_kind](checkpoint=modelname) + size = os.path.getsize(modelname) + safe_to = core.SafeToGPU(size) + + # Unless user explicitly wants to use CPU, we use GPU + device = comfy.model_management.get_torch_device() if device_mode == "Prefer GPU" else "CPU" + + if device_mode == "Prefer GPU": + safe_to.to_device(sam, device) + + is_auto_mode = device_mode == "AUTO" + + sam_obj = core.SAMWrapper(sam, is_auto_mode=is_auto_mode, safe_to_gpu=safe_to) + sam.sam_wrapper = sam_obj + + print(f"Loads SAM model: {modelname} (device:{device_mode})") + return (sam, ) + + +class ONNXDetectorForEach: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "onnx_detector": ("ONNX_DETECTOR",), + "image": ("IMAGE",), + "threshold": ("FLOAT", {"default": 0.8, "min": 0.0, "max": 1.0, "step": 0.01}), + "dilation": ("INT", {"default": 10, "min": -512, "max": 512, "step": 1}), + "crop_factor": ("FLOAT", {"default": 1.0, "min": 0.5, "max": 100, "step": 0.1}), + "drop_size": ("INT", {"min": 1, "max": MAX_RESOLUTION, "step": 1, "default": 10}), + } + } + + RETURN_TYPES = ("SEGS", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detector" + + OUTPUT_NODE = True + + def doit(self, onnx_detector, image, threshold, dilation, crop_factor, drop_size): + segs = onnx_detector.detect(image, threshold, dilation, crop_factor, drop_size) + return (segs, ) + + +class DetailerForEach: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "image": ("IMAGE", ), + "segs": ("SEGS", ), + "model": ("MODEL",), + "clip": ("CLIP",), + "vae": ("VAE",), + "guide_size": ("FLOAT", {"default": 512, "min": 64, "max": nodes.MAX_RESOLUTION, "step": 8}), + "guide_size_for": ("BOOLEAN", {"default": True, "label_on": "bbox", "label_off": "crop_region"}), + "max_size": ("FLOAT", {"default": 1024, "min": 64, "max": nodes.MAX_RESOLUTION, "step": 8}), + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000}), + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS,), + "scheduler": (core.SCHEDULERS,), + "positive": ("CONDITIONING",), + "negative": ("CONDITIONING",), + "denoise": ("FLOAT", {"default": 0.5, "min": 0.0001, "max": 1.0, "step": 0.01}), + "feather": ("INT", {"default": 5, "min": 0, "max": 100, "step": 1}), + "noise_mask": ("BOOLEAN", {"default": True, "label_on": "enabled", "label_off": "disabled"}), + "force_inpaint": ("BOOLEAN", {"default": True, "label_on": "enabled", "label_off": "disabled"}), + "wildcard": ("STRING", {"multiline": True, "dynamicPrompts": False}), + + "cycle": ("INT", {"default": 1, "min": 1, "max": 10, "step": 1}), + }, + "optional": { + "detailer_hook": ("DETAILER_HOOK",), + "inpaint_model": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + "noise_mask_feather": ("INT", {"default": 20, "min": 0, "max": 100, "step": 1}), + "scheduler_func_opt": ("SCHEDULER_FUNC",), + } + } + + RETURN_TYPES = ("IMAGE", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detailer" + + @staticmethod + def do_detail(image, segs, model, clip, vae, guide_size, guide_size_for_bbox, max_size, seed, steps, cfg, sampler_name, scheduler, + positive, negative, denoise, feather, noise_mask, force_inpaint, wildcard_opt=None, detailer_hook=None, + refiner_ratio=None, refiner_model=None, refiner_clip=None, refiner_positive=None, refiner_negative=None, + cycle=1, inpaint_model=False, noise_mask_feather=0, scheduler_func_opt=None): + + if len(image) > 1: + raise Exception('[Impact Pack] ERROR: DetailerForEach does not allow image batches.\nPlease refer to https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/batching-detailer.md for more information.') + + image = image.clone() + enhanced_alpha_list = [] + enhanced_list = [] + cropped_list = [] + cnet_pil_list = [] + + segs = core.segs_scale_match(segs, image.shape) + new_segs = [] + + wildcard_concat_mode = None + if wildcard_opt is not None: + if wildcard_opt.startswith('[CONCAT]'): + wildcard_concat_mode = 'concat' + wildcard_opt = wildcard_opt[8:] + wmode, wildcard_chooser = wildcards.process_wildcard_for_segs(wildcard_opt) + else: + wmode, wildcard_chooser = None, None + + if wmode in ['ASC', 'DSC', 'ASC-SIZE', 'DSC-SIZE']: + if wmode == 'ASC': + ordered_segs = sorted(segs[1], key=lambda x: (x.bbox[0], x.bbox[1])) + elif wmode == 'DSC': + ordered_segs = sorted(segs[1], key=lambda x: (x.bbox[0], x.bbox[1]), reverse=True) + elif wmode == 'ASC-SIZE': + ordered_segs = sorted(segs[1], key=lambda x: (x.bbox[2]-x.bbox[0]) * (x.bbox[3]-x.bbox[1])) + + else: # wmode == 'DSC-SIZE' + ordered_segs = sorted(segs[1], key=lambda x: (x.bbox[2]-x.bbox[0]) * (x.bbox[3]-x.bbox[1]), reverse=True) + else: + ordered_segs = segs[1] + + if noise_mask_feather > 0 and 'denoise_mask_function' not in model.model_options: + model = nodes_differential_diffusion.DifferentialDiffusion().apply(model)[0] + + for i, seg in enumerate(ordered_segs): + cropped_image = crop_ndarray4(image.cpu().numpy(), seg.crop_region) # Never use seg.cropped_image to handle overlapping area + cropped_image = to_tensor(cropped_image) + mask = to_tensor(seg.cropped_mask) + mask = tensor_gaussian_blur_mask(mask, feather) + + is_mask_all_zeros = (seg.cropped_mask == 0).all().item() + if is_mask_all_zeros: + print(f"Detailer: segment skip [empty mask]") + continue + + if noise_mask: + cropped_mask = seg.cropped_mask + else: + cropped_mask = None + + if wildcard_chooser is not None and wmode != "LAB": + seg_seed, wildcard_item = wildcard_chooser.get(seg) + elif wildcard_chooser is not None and wmode == "LAB": + seg_seed, wildcard_item = None, wildcard_chooser.get(seg) + else: + seg_seed, wildcard_item = None, None + + seg_seed = seed + i if seg_seed is None else seg_seed + + cropped_positive = [ + [condition, { + k: core.crop_condition_mask(v, image, seg.crop_region) if k == "mask" else v + for k, v in details.items() + }] + for condition, details in positive + ] + + if not isinstance(negative, str): + cropped_negative = [ + [condition, { + k: core.crop_condition_mask(v, image, seg.crop_region) if k == "mask" else v + for k, v in details.items() + }] + for condition, details in negative + ] + else: + # Negative Conditioning is placeholder such as FLUX.1 + cropped_negative = negative + + if wildcard_item and wildcard_item.strip() == '[SKIP]': + continue + + if wildcard_item and wildcard_item.strip() == '[STOP]': + break + + orig_cropped_image = cropped_image.clone() + enhanced_image, cnet_pils = core.enhance_detail(cropped_image, model, clip, vae, guide_size, guide_size_for_bbox, max_size, + seg.bbox, seg_seed, steps, cfg, sampler_name, scheduler, + cropped_positive, cropped_negative, denoise, cropped_mask, force_inpaint, + wildcard_opt=wildcard_item, wildcard_opt_concat_mode=wildcard_concat_mode, + detailer_hook=detailer_hook, + refiner_ratio=refiner_ratio, refiner_model=refiner_model, + refiner_clip=refiner_clip, refiner_positive=refiner_positive, + refiner_negative=refiner_negative, control_net_wrapper=seg.control_net_wrapper, + cycle=cycle, inpaint_model=inpaint_model, noise_mask_feather=noise_mask_feather, + scheduler_func=scheduler_func_opt) + + if cnet_pils is not None: + cnet_pil_list.extend(cnet_pils) + + if not (enhanced_image is None): + # don't latent composite-> converting to latent caused poor quality + # use image paste + image = image.cpu() + enhanced_image = enhanced_image.cpu() + tensor_paste(image, enhanced_image, (seg.crop_region[0], seg.crop_region[1]), mask) # this code affecting to `cropped_image`. + enhanced_list.append(enhanced_image) + + if detailer_hook is not None: + image = detailer_hook.post_paste(image) + + if not (enhanced_image is None): + # Convert enhanced_pil_alpha to RGBA mode + enhanced_image_alpha = tensor_convert_rgba(enhanced_image) + new_seg_image = enhanced_image.numpy() # alpha should not be applied to seg_image + + # Apply the mask + mask = tensor_resize(mask, *tensor_get_size(enhanced_image)) + tensor_putalpha(enhanced_image_alpha, mask) + enhanced_alpha_list.append(enhanced_image_alpha) + else: + new_seg_image = None + + cropped_list.append(orig_cropped_image) # NOTE: Don't use `cropped_image` + + new_seg = SEG(new_seg_image, seg.cropped_mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, seg.control_net_wrapper) + new_segs.append(new_seg) + + image_tensor = tensor_convert_rgb(image) + + cropped_list.sort(key=lambda x: x.shape, reverse=True) + enhanced_list.sort(key=lambda x: x.shape, reverse=True) + enhanced_alpha_list.sort(key=lambda x: x.shape, reverse=True) + + return image_tensor, cropped_list, enhanced_list, enhanced_alpha_list, cnet_pil_list, (segs[0], new_segs) + + def doit(self, image, segs, model, clip, vae, guide_size, guide_size_for, max_size, seed, steps, cfg, sampler_name, + scheduler, positive, negative, denoise, feather, noise_mask, force_inpaint, wildcard, cycle=1, + detailer_hook=None, inpaint_model=False, noise_mask_feather=0, scheduler_func_opt=None): + + enhanced_img, *_ = \ + DetailerForEach.do_detail(image, segs, model, clip, vae, guide_size, guide_size_for, max_size, seed, steps, + cfg, sampler_name, scheduler, positive, negative, denoise, feather, noise_mask, + force_inpaint, wildcard, detailer_hook, + cycle=cycle, inpaint_model=inpaint_model, noise_mask_feather=noise_mask_feather, scheduler_func_opt=scheduler_func_opt) + + return (enhanced_img, ) + + +class DetailerForEachPipe: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "image": ("IMAGE", ), + "segs": ("SEGS", ), + "guide_size": ("FLOAT", {"default": 512, "min": 64, "max": nodes.MAX_RESOLUTION, "step": 8}), + "guide_size_for": ("BOOLEAN", {"default": True, "label_on": "bbox", "label_off": "crop_region"}), + "max_size": ("FLOAT", {"default": 1024, "min": 64, "max": nodes.MAX_RESOLUTION, "step": 8}), + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000}), + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS,), + "scheduler": (core.SCHEDULERS,), + "denoise": ("FLOAT", {"default": 0.5, "min": 0.0001, "max": 1.0, "step": 0.01}), + "feather": ("INT", {"default": 5, "min": 0, "max": 100, "step": 1}), + "noise_mask": ("BOOLEAN", {"default": True, "label_on": "enabled", "label_off": "disabled"}), + "force_inpaint": ("BOOLEAN", {"default": True, "label_on": "enabled", "label_off": "disabled"}), + "basic_pipe": ("BASIC_PIPE", ), + "wildcard": ("STRING", {"multiline": True, "dynamicPrompts": False}), + "refiner_ratio": ("FLOAT", {"default": 0.2, "min": 0.0, "max": 1.0}), + + "cycle": ("INT", {"default": 1, "min": 1, "max": 10, "step": 1}), + }, + "optional": { + "detailer_hook": ("DETAILER_HOOK",), + "refiner_basic_pipe_opt": ("BASIC_PIPE",), + "inpaint_model": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + "noise_mask_feather": ("INT", {"default": 20, "min": 0, "max": 100, "step": 1}), + "scheduler_func_opt": ("SCHEDULER_FUNC",), + } + } + + RETURN_TYPES = ("IMAGE", "SEGS", "BASIC_PIPE", "IMAGE") + RETURN_NAMES = ("image", "segs", "basic_pipe", "cnet_images") + OUTPUT_IS_LIST = (False, False, False, True) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detailer" + + def doit(self, image, segs, guide_size, guide_size_for, max_size, seed, steps, cfg, sampler_name, scheduler, + denoise, feather, noise_mask, force_inpaint, basic_pipe, wildcard, + refiner_ratio=None, detailer_hook=None, refiner_basic_pipe_opt=None, + cycle=1, inpaint_model=False, noise_mask_feather=0, scheduler_func_opt=None): + + if len(image) > 1: + raise Exception('[Impact Pack] ERROR: DetailerForEach does not allow image batches.\nPlease refer to https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/batching-detailer.md for more information.') + + model, clip, vae, positive, negative = basic_pipe + + if refiner_basic_pipe_opt is None: + refiner_model, refiner_clip, refiner_positive, refiner_negative = None, None, None, None + else: + refiner_model, refiner_clip, _, refiner_positive, refiner_negative = refiner_basic_pipe_opt + + enhanced_img, cropped, cropped_enhanced, cropped_enhanced_alpha, cnet_pil_list, new_segs = \ + DetailerForEach.do_detail(image, segs, model, clip, vae, guide_size, guide_size_for, max_size, seed, steps, cfg, + sampler_name, scheduler, positive, negative, denoise, feather, noise_mask, + force_inpaint, wildcard, detailer_hook, + refiner_ratio=refiner_ratio, refiner_model=refiner_model, + refiner_clip=refiner_clip, refiner_positive=refiner_positive, refiner_negative=refiner_negative, + cycle=cycle, inpaint_model=inpaint_model, noise_mask_feather=noise_mask_feather, scheduler_func_opt=scheduler_func_opt) + + # set fallback image + if len(cnet_pil_list) == 0: + cnet_pil_list = [empty_pil_tensor()] + + return enhanced_img, new_segs, basic_pipe, cnet_pil_list + + +class FaceDetailer: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "image": ("IMAGE", ), + "model": ("MODEL",), + "clip": ("CLIP",), + "vae": ("VAE",), + "guide_size": ("FLOAT", {"default": 512, "min": 64, "max": nodes.MAX_RESOLUTION, "step": 8}), + "guide_size_for": ("BOOLEAN", {"default": True, "label_on": "bbox", "label_off": "crop_region"}), + "max_size": ("FLOAT", {"default": 1024, "min": 64, "max": nodes.MAX_RESOLUTION, "step": 8}), + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000}), + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS,), + "scheduler": (core.SCHEDULERS,), + "positive": ("CONDITIONING",), + "negative": ("CONDITIONING",), + "denoise": ("FLOAT", {"default": 0.5, "min": 0.0001, "max": 1.0, "step": 0.01}), + "feather": ("INT", {"default": 5, "min": 0, "max": 100, "step": 1}), + "noise_mask": ("BOOLEAN", {"default": True, "label_on": "enabled", "label_off": "disabled"}), + "force_inpaint": ("BOOLEAN", {"default": True, "label_on": "enabled", "label_off": "disabled"}), + + "bbox_threshold": ("FLOAT", {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01}), + "bbox_dilation": ("INT", {"default": 10, "min": -512, "max": 512, "step": 1}), + "bbox_crop_factor": ("FLOAT", {"default": 3.0, "min": 1.0, "max": 10, "step": 0.1}), + + "sam_detection_hint": (["center-1", "horizontal-2", "vertical-2", "rect-4", "diamond-4", "mask-area", "mask-points", "mask-point-bbox", "none"],), + "sam_dilation": ("INT", {"default": 0, "min": -512, "max": 512, "step": 1}), + "sam_threshold": ("FLOAT", {"default": 0.93, "min": 0.0, "max": 1.0, "step": 0.01}), + "sam_bbox_expansion": ("INT", {"default": 0, "min": 0, "max": 1000, "step": 1}), + "sam_mask_hint_threshold": ("FLOAT", {"default": 0.7, "min": 0.0, "max": 1.0, "step": 0.01}), + "sam_mask_hint_use_negative": (["False", "Small", "Outter"],), + + "drop_size": ("INT", {"min": 1, "max": MAX_RESOLUTION, "step": 1, "default": 10}), + + "bbox_detector": ("BBOX_DETECTOR", ), + "wildcard": ("STRING", {"multiline": True, "dynamicPrompts": False}), + + "cycle": ("INT", {"default": 1, "min": 1, "max": 10, "step": 1}), + }, + "optional": { + "sam_model_opt": ("SAM_MODEL", ), + "segm_detector_opt": ("SEGM_DETECTOR", ), + "detailer_hook": ("DETAILER_HOOK",), + "inpaint_model": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + "noise_mask_feather": ("INT", {"default": 20, "min": 0, "max": 100, "step": 1}), + "scheduler_func_opt": ("SCHEDULER_FUNC",), + }} + + RETURN_TYPES = ("IMAGE", "IMAGE", "IMAGE", "MASK", "DETAILER_PIPE", "IMAGE") + RETURN_NAMES = ("image", "cropped_refined", "cropped_enhanced_alpha", "mask", "detailer_pipe", "cnet_images") + OUTPUT_IS_LIST = (False, True, True, False, False, True) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Simple" + + @staticmethod + def enhance_face(image, model, clip, vae, guide_size, guide_size_for_bbox, max_size, seed, steps, cfg, sampler_name, scheduler, + positive, negative, denoise, feather, noise_mask, force_inpaint, + bbox_threshold, bbox_dilation, bbox_crop_factor, + sam_detection_hint, sam_dilation, sam_threshold, sam_bbox_expansion, sam_mask_hint_threshold, + sam_mask_hint_use_negative, drop_size, + bbox_detector, segm_detector=None, sam_model_opt=None, wildcard_opt=None, detailer_hook=None, + refiner_ratio=None, refiner_model=None, refiner_clip=None, refiner_positive=None, refiner_negative=None, cycle=1, + inpaint_model=False, noise_mask_feather=0, scheduler_func_opt=None): + + # make default prompt as 'face' if empty prompt for CLIPSeg + bbox_detector.setAux('face') + segs = bbox_detector.detect(image, bbox_threshold, bbox_dilation, bbox_crop_factor, drop_size, detailer_hook=detailer_hook) + bbox_detector.setAux(None) + + # bbox + sam combination + if sam_model_opt is not None: + sam_mask = core.make_sam_mask(sam_model_opt, segs, image, sam_detection_hint, sam_dilation, + sam_threshold, sam_bbox_expansion, sam_mask_hint_threshold, + sam_mask_hint_use_negative, ) + segs = core.segs_bitwise_and_mask(segs, sam_mask) + + elif segm_detector is not None: + segm_segs = segm_detector.detect(image, bbox_threshold, bbox_dilation, bbox_crop_factor, drop_size) + + if (hasattr(segm_detector, 'override_bbox_by_segm') and segm_detector.override_bbox_by_segm and + not (detailer_hook is not None and not hasattr(detailer_hook, 'override_bbox_by_segm'))): + segs = segm_segs + else: + segm_mask = core.segs_to_combined_mask(segm_segs) + segs = core.segs_bitwise_and_mask(segs, segm_mask) + + if len(segs[1]) > 0: + enhanced_img, _, cropped_enhanced, cropped_enhanced_alpha, cnet_pil_list, new_segs = \ + DetailerForEach.do_detail(image, segs, model, clip, vae, guide_size, guide_size_for_bbox, max_size, seed, steps, cfg, + sampler_name, scheduler, positive, negative, denoise, feather, noise_mask, + force_inpaint, wildcard_opt, detailer_hook, + refiner_ratio=refiner_ratio, refiner_model=refiner_model, + refiner_clip=refiner_clip, refiner_positive=refiner_positive, + refiner_negative=refiner_negative, + cycle=cycle, inpaint_model=inpaint_model, noise_mask_feather=noise_mask_feather, scheduler_func_opt=scheduler_func_opt) + else: + enhanced_img = image + cropped_enhanced = [] + cropped_enhanced_alpha = [] + cnet_pil_list = [] + + # Mask Generator + mask = core.segs_to_combined_mask(segs) + + if len(cropped_enhanced) == 0: + cropped_enhanced = [empty_pil_tensor()] + + if len(cropped_enhanced_alpha) == 0: + cropped_enhanced_alpha = [empty_pil_tensor()] + + if len(cnet_pil_list) == 0: + cnet_pil_list = [empty_pil_tensor()] + + return enhanced_img, cropped_enhanced, cropped_enhanced_alpha, mask, cnet_pil_list + + def doit(self, image, model, clip, vae, guide_size, guide_size_for, max_size, seed, steps, cfg, sampler_name, scheduler, + positive, negative, denoise, feather, noise_mask, force_inpaint, + bbox_threshold, bbox_dilation, bbox_crop_factor, + sam_detection_hint, sam_dilation, sam_threshold, sam_bbox_expansion, sam_mask_hint_threshold, + sam_mask_hint_use_negative, drop_size, bbox_detector, wildcard, cycle=1, + sam_model_opt=None, segm_detector_opt=None, detailer_hook=None, inpaint_model=False, noise_mask_feather=0, scheduler_func_opt=None): + + result_img = None + result_mask = None + result_cropped_enhanced = [] + result_cropped_enhanced_alpha = [] + result_cnet_images = [] + + if len(image) > 1: + print(f"[Impact Pack] WARN: FaceDetailer is not a node designed for video detailing. If you intend to perform video detailing, please use Detailer For AnimateDiff.") + + for i, single_image in enumerate(image): + enhanced_img, cropped_enhanced, cropped_enhanced_alpha, mask, cnet_pil_list = FaceDetailer.enhance_face( + single_image.unsqueeze(0), model, clip, vae, guide_size, guide_size_for, max_size, seed + i, steps, cfg, sampler_name, scheduler, + positive, negative, denoise, feather, noise_mask, force_inpaint, + bbox_threshold, bbox_dilation, bbox_crop_factor, + sam_detection_hint, sam_dilation, sam_threshold, sam_bbox_expansion, sam_mask_hint_threshold, + sam_mask_hint_use_negative, drop_size, bbox_detector, segm_detector_opt, sam_model_opt, wildcard, detailer_hook, + cycle=cycle, inpaint_model=inpaint_model, noise_mask_feather=noise_mask_feather, scheduler_func_opt=scheduler_func_opt) + + result_img = torch.cat((result_img, enhanced_img), dim=0) if result_img is not None else enhanced_img + result_mask = torch.cat((result_mask, mask), dim=0) if result_mask is not None else mask + result_cropped_enhanced.extend(cropped_enhanced) + result_cropped_enhanced_alpha.extend(cropped_enhanced_alpha) + result_cnet_images.extend(cnet_pil_list) + + pipe = (model, clip, vae, positive, negative, wildcard, bbox_detector, segm_detector_opt, sam_model_opt, detailer_hook, None, None, None, None) + return result_img, result_cropped_enhanced, result_cropped_enhanced_alpha, result_mask, pipe, result_cnet_images + + +class LatentPixelScale: + upscale_methods = ["nearest-exact", "bilinear", "lanczos", "area"] + + @classmethod + def INPUT_TYPES(s): + return {"required": { + "samples": ("LATENT", ), + "scale_method": (s.upscale_methods,), + "scale_factor": ("FLOAT", {"default": 1.5, "min": 0.1, "max": 10000, "step": 0.1}), + "vae": ("VAE", ), + "use_tiled_vae": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + }, + "optional": { + "upscale_model_opt": ("UPSCALE_MODEL", ), + } + } + + RETURN_TYPES = ("LATENT", "IMAGE") + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Upscale" + + def doit(self, samples, scale_method, scale_factor, vae, use_tiled_vae, upscale_model_opt=None): + if upscale_model_opt is None: + latimg = core.latent_upscale_on_pixel_space2(samples, scale_method, scale_factor, vae, use_tile=use_tiled_vae) + else: + latimg = core.latent_upscale_on_pixel_space_with_model2(samples, scale_method, upscale_model_opt, scale_factor, vae, use_tile=use_tiled_vae) + return latimg + + +class NoiseInjectionDetailerHookProvider: + schedules = ["skip_start", "from_start"] + + @classmethod + def INPUT_TYPES(s): + return {"required": { + "schedule_for_cycle": (s.schedules,), + "source": (["CPU", "GPU"],), + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "start_strength": ("FLOAT", {"default": 2.0, "min": 0.0, "max": 200.0, "step": 0.01}), + "end_strength": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 200.0, "step": 0.01}), + }, + } + + RETURN_TYPES = ("DETAILER_HOOK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detailer" + + def doit(self, schedule_for_cycle, source, seed, start_strength, end_strength): + try: + hook = hooks.InjectNoiseHookForDetailer(source, seed, start_strength, end_strength, + from_start=('from_start' in schedule_for_cycle)) + return (hook, ) + except Exception as e: + print("[ERROR] NoiseInjectionDetailerHookProvider: 'ComfyUI Noise' custom node isn't installed. You must install 'BlenderNeko/ComfyUI Noise' extension to use this node.") + print(f"\t{e}") + pass + + +# class CustomNoiseDetailerHookProvider: +# @classmethod +# def INPUT_TYPES(s): +# return {"required": { +# "noise": ("NOISE",)}, +# } +# +# RETURN_TYPES = ("DETAILER_HOOK",) +# FUNCTION = "doit" +# +# CATEGORY = "ImpactPack/Detailer" +# +# def doit(self, noise): +# hook = hooks.CustomNoiseDetailerHookProvider(noise) +# return (hook, ) + + +class VariationNoiseDetailerHookProvider: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "strength": ("FLOAT", {"default": 0.0, "min": 0.0, "max": 1.0, "step": 0.01})} + } + + RETURN_TYPES = ("DETAILER_HOOK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detailer" + + def doit(self, seed, strength): + hook = hooks.VariationNoiseDetailerHookProvider(seed, strength) + return (hook, ) + + +class UnsamplerDetailerHookProvider: + schedules = ["skip_start", "from_start"] + + @classmethod + def INPUT_TYPES(s): + return {"required": + {"model": ("MODEL",), + "steps": ("INT", {"default": 25, "min": 1, "max": 10000}), + "start_end_at_step": ("INT", {"default": 21, "min": 0, "max": 10000}), + "end_end_at_step": ("INT", {"default": 24, "min": 0, "max": 10000}), + "cfg": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 100.0}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS, ), + "scheduler": (comfy.samplers.KSampler.SCHEDULERS, ), + "normalize": (["disable", "enable"], ), + "positive": ("CONDITIONING", ), + "negative": ("CONDITIONING", ), + "schedule_for_cycle": (s.schedules,), + }} + + RETURN_TYPES = ("DETAILER_HOOK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detailer" + + def doit(self, model, steps, start_end_at_step, end_end_at_step, cfg, sampler_name, + scheduler, normalize, positive, negative, schedule_for_cycle): + try: + hook = hooks.UnsamplerDetailerHook(model, steps, start_end_at_step, end_end_at_step, cfg, sampler_name, + scheduler, normalize, positive, negative, + from_start=('from_start' in schedule_for_cycle)) + + return (hook, ) + except Exception as e: + print("[ERROR] UnsamplerDetailerHookProvider: 'ComfyUI Noise' custom node isn't installed. You must install 'BlenderNeko/ComfyUI Noise' extension to use this node.") + print(f"\t{e}") + pass + + +class DenoiseSchedulerDetailerHookProvider: + schedules = ["simple"] + + @classmethod + def INPUT_TYPES(s): + return {"required": { + "schedule_for_cycle": (s.schedules,), + "target_denoise": ("FLOAT", {"default": 0.3, "min": 0.0, "max": 1.0, "step": 0.01}), + }, + } + + RETURN_TYPES = ("DETAILER_HOOK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detailer" + + def doit(self, schedule_for_cycle, target_denoise): + hook = hooks.SimpleDetailerDenoiseSchedulerHook(target_denoise) + return (hook, ) + + +class CoreMLDetailerHookProvider: + @classmethod + def INPUT_TYPES(s): + return {"required": {"mode": (["512x512", "768x768", "512x768", "768x512"], )}, } + + RETURN_TYPES = ("DETAILER_HOOK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detailer" + + def doit(self, mode): + hook = hooks.CoreMLHook(mode) + return (hook, ) + + +class CfgScheduleHookProvider: + schedules = ["simple"] + + @classmethod + def INPUT_TYPES(s): + return {"required": { + "schedule_for_iteration": (s.schedules,), + "target_cfg": ("FLOAT", {"default": 3.0, "min": 0.0, "max": 100.0}), + }, + } + + RETURN_TYPES = ("PK_HOOK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Upscale" + + def doit(self, schedule_for_iteration, target_cfg): + hook = None + if schedule_for_iteration == "simple": + hook = hooks.SimpleCfgScheduleHook(target_cfg) + + return (hook, ) + + +class UnsamplerHookProvider: + schedules = ["simple"] + + @classmethod + def INPUT_TYPES(s): + return {"required": + {"model": ("MODEL",), + "steps": ("INT", {"default": 25, "min": 1, "max": 10000}), + "start_end_at_step": ("INT", {"default": 21, "min": 0, "max": 10000}), + "end_end_at_step": ("INT", {"default": 24, "min": 0, "max": 10000}), + "cfg": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 100.0}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS, ), + "scheduler": (comfy.samplers.KSampler.SCHEDULERS, ), + "normalize": (["disable", "enable"], ), + "positive": ("CONDITIONING", ), + "negative": ("CONDITIONING", ), + "schedule_for_iteration": (s.schedules,), + }} + + RETURN_TYPES = ("PK_HOOK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Upscale" + + def doit(self, model, steps, start_end_at_step, end_end_at_step, cfg, sampler_name, + scheduler, normalize, positive, negative, schedule_for_iteration): + try: + hook = None + if schedule_for_iteration == "simple": + hook = hooks.UnsamplerHook(model, steps, start_end_at_step, end_end_at_step, cfg, sampler_name, + scheduler, normalize, positive, negative) + + return (hook, ) + except Exception as e: + print("[ERROR] UnsamplerHookProvider: 'ComfyUI Noise' custom node isn't installed. You must install 'BlenderNeko/ComfyUI Noise' extension to use this node.") + print(f"\t{e}") + pass + + +class NoiseInjectionHookProvider: + schedules = ["simple"] + + @classmethod + def INPUT_TYPES(s): + return {"required": { + "schedule_for_iteration": (s.schedules,), + "source": (["CPU", "GPU"],), + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "start_strength": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 200.0, "step": 0.01}), + "end_strength": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 200.0, "step": 0.01}), + }, + } + + RETURN_TYPES = ("PK_HOOK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Upscale" + + def doit(self, schedule_for_iteration, source, seed, start_strength, end_strength): + try: + hook = None + if schedule_for_iteration == "simple": + hook = hooks.InjectNoiseHook(source, seed, start_strength, end_strength) + + return (hook, ) + except Exception as e: + print("[ERROR] NoiseInjectionHookProvider: 'ComfyUI Noise' custom node isn't installed. You must install 'BlenderNeko/ComfyUI Noise' extension to use this node.") + print(f"\t{e}") + pass + + +class DenoiseScheduleHookProvider: + schedules = ["simple"] + + @classmethod + def INPUT_TYPES(s): + return {"required": { + "schedule_for_iteration": (s.schedules,), + "target_denoise": ("FLOAT", {"default": 0.2, "min": 0.0, "max": 1.0, "step": 0.01}), + }, + } + + RETURN_TYPES = ("PK_HOOK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Upscale" + + def doit(self, schedule_for_iteration, target_denoise): + hook = None + if schedule_for_iteration == "simple": + hook = hooks.SimpleDenoiseScheduleHook(target_denoise) + + return (hook, ) + + +class StepsScheduleHookProvider: + schedules = ["simple"] + + @classmethod + def INPUT_TYPES(s): + return {"required": { + "schedule_for_iteration": (s.schedules,), + "target_steps": ("INT", {"default": 20, "min": 1, "max": 10000}), + }, + } + + RETURN_TYPES = ("PK_HOOK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Upscale" + + def doit(self, schedule_for_iteration, target_steps): + hook = None + if schedule_for_iteration == "simple": + hook = hooks.SimpleStepsScheduleHook(target_steps) + + return (hook, ) + + +class DetailerHookCombine: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "hook1": ("DETAILER_HOOK",), + "hook2": ("DETAILER_HOOK",), + }, + } + + RETURN_TYPES = ("DETAILER_HOOK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Upscale" + + def doit(self, hook1, hook2): + hook = hooks.DetailerHookCombine(hook1, hook2) + return (hook, ) + + +class PixelKSampleHookCombine: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "hook1": ("PK_HOOK",), + "hook2": ("PK_HOOK",), + }, + } + + RETURN_TYPES = ("PK_HOOK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Upscale" + + def doit(self, hook1, hook2): + hook = hooks.PixelKSampleHookCombine(hook1, hook2) + return (hook, ) + + +class PixelTiledKSampleUpscalerProvider: + upscale_methods = ["nearest-exact", "bilinear", "lanczos", "area"] + + @classmethod + def INPUT_TYPES(s): + return {"required": { + "scale_method": (s.upscale_methods,), + "model": ("MODEL",), + "vae": ("VAE",), + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000}), + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS, ), + "scheduler": (comfy.samplers.KSampler.SCHEDULERS, ), + "positive": ("CONDITIONING", ), + "negative": ("CONDITIONING", ), + "denoise": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01}), + "tile_width": ("INT", {"default": 512, "min": 320, "max": MAX_RESOLUTION, "step": 64}), + "tile_height": ("INT", {"default": 512, "min": 320, "max": MAX_RESOLUTION, "step": 64}), + "tiling_strategy": (["random", "padded", 'simple'], ), + }, + "optional": { + "upscale_model_opt": ("UPSCALE_MODEL", ), + "pk_hook_opt": ("PK_HOOK", ), + "tile_cnet_opt": ("CONTROL_NET", ), + "tile_cnet_strength": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01}), + } + } + + RETURN_TYPES = ("UPSCALER",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Upscale" + + def doit(self, scale_method, model, vae, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise, tile_width, tile_height, tiling_strategy, upscale_model_opt=None, + pk_hook_opt=None, tile_cnet_opt=None, tile_cnet_strength=1.0): + if "BNK_TiledKSampler" in nodes.NODE_CLASS_MAPPINGS: + upscaler = core.PixelTiledKSampleUpscaler(scale_method, model, vae, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise, + tile_width, tile_height, tiling_strategy, upscale_model_opt, pk_hook_opt, tile_cnet_opt, + tile_size=max(tile_width, tile_height), tile_cnet_strength=tile_cnet_strength) + return (upscaler, ) + else: + utils.try_install_custom_node('https://github.com/BlenderNeko/ComfyUI_TiledKSampler', + "To use 'PixelTiledKSampleUpscalerProvider' node, 'BlenderNeko/ComfyUI_TiledKSampler' extension is required.") + + raise Exception("[ERROR] PixelTiledKSampleUpscalerProvider: ComfyUI_TiledKSampler custom node isn't installed. You must install BlenderNeko/ComfyUI_TiledKSampler extension to use this node.") + + +class PixelTiledKSampleUpscalerProviderPipe: + upscale_methods = ["nearest-exact", "bilinear", "lanczos", "area"] + + @classmethod + def INPUT_TYPES(s): + return {"required": { + "scale_method": (s.upscale_methods,), + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000}), + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS, ), + "scheduler": (comfy.samplers.KSampler.SCHEDULERS, ), + "denoise": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01}), + "tile_width": ("INT", {"default": 512, "min": 320, "max": MAX_RESOLUTION, "step": 64}), + "tile_height": ("INT", {"default": 512, "min": 320, "max": MAX_RESOLUTION, "step": 64}), + "tiling_strategy": (["random", "padded", 'simple'], ), + "basic_pipe": ("BASIC_PIPE",) + }, + "optional": { + "upscale_model_opt": ("UPSCALE_MODEL", ), + "pk_hook_opt": ("PK_HOOK", ), + "tile_cnet_opt": ("CONTROL_NET", ), + "tile_cnet_strength": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01}), + } + } + + RETURN_TYPES = ("UPSCALER",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Upscale" + + def doit(self, scale_method, seed, steps, cfg, sampler_name, scheduler, denoise, tile_width, tile_height, tiling_strategy, basic_pipe, upscale_model_opt=None, pk_hook_opt=None, + tile_cnet_opt=None, tile_cnet_strength=1.0): + if "BNK_TiledKSampler" in nodes.NODE_CLASS_MAPPINGS: + model, _, vae, positive, negative = basic_pipe + upscaler = core.PixelTiledKSampleUpscaler(scale_method, model, vae, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise, + tile_width, tile_height, tiling_strategy, upscale_model_opt, pk_hook_opt, tile_cnet_opt, + tile_size=max(tile_width, tile_height), tile_cnet_strength=tile_cnet_strength) + return (upscaler, ) + else: + print("[ERROR] PixelTiledKSampleUpscalerProviderPipe: ComfyUI_TiledKSampler custom node isn't installed. You must install BlenderNeko/ComfyUI_TiledKSampler extension to use this node.") + + +class PixelKSampleUpscalerProvider: + upscale_methods = ["nearest-exact", "bilinear", "lanczos", "area"] + + @classmethod + def INPUT_TYPES(s): + return {"required": { + "scale_method": (s.upscale_methods,), + "model": ("MODEL",), + "vae": ("VAE",), + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000}), + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS, ), + "scheduler": (core.SCHEDULERS, ), + "positive": ("CONDITIONING", ), + "negative": ("CONDITIONING", ), + "denoise": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01}), + "use_tiled_vae": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + "tile_size": ("INT", {"default": 512, "min": 320, "max": 4096, "step": 64}), + }, + "optional": { + "upscale_model_opt": ("UPSCALE_MODEL", ), + "pk_hook_opt": ("PK_HOOK", ), + "scheduler_func_opt": ("SCHEDULER_FUNC",), + } + } + + RETURN_TYPES = ("UPSCALER",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Upscale" + + def doit(self, scale_method, model, vae, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise, + use_tiled_vae, upscale_model_opt=None, pk_hook_opt=None, tile_size=512, scheduler_func_opt=None): + upscaler = core.PixelKSampleUpscaler(scale_method, model, vae, seed, steps, cfg, sampler_name, scheduler, + positive, negative, denoise, use_tiled_vae, upscale_model_opt, pk_hook_opt, + tile_size=tile_size, scheduler_func=scheduler_func_opt) + return (upscaler, ) + + +class PixelKSampleUpscalerProviderPipe(PixelKSampleUpscalerProvider): + upscale_methods = ["nearest-exact", "bilinear", "lanczos", "area"] + + @classmethod + def INPUT_TYPES(s): + return {"required": { + "scale_method": (s.upscale_methods,), + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000}), + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS, ), + "scheduler": (core.SCHEDULERS, ), + "denoise": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01}), + "use_tiled_vae": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + "basic_pipe": ("BASIC_PIPE",), + "tile_size": ("INT", {"default": 512, "min": 320, "max": 4096, "step": 64}), + }, + "optional": { + "upscale_model_opt": ("UPSCALE_MODEL", ), + "pk_hook_opt": ("PK_HOOK", ), + "scheduler_func_opt": ("SCHEDULER_FUNC",), + "tile_cnet_opt": ("CONTROL_NET", ), + "tile_cnet_strength": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01}), + } + } + + RETURN_TYPES = ("UPSCALER",) + FUNCTION = "doit_pipe" + + CATEGORY = "ImpactPack/Upscale" + + def doit_pipe(self, scale_method, seed, steps, cfg, sampler_name, scheduler, denoise, + use_tiled_vae, basic_pipe, upscale_model_opt=None, pk_hook_opt=None, + tile_size=512, scheduler_func_opt=None, tile_cnet_opt=None, tile_cnet_strength=1.0): + model, _, vae, positive, negative = basic_pipe + upscaler = core.PixelKSampleUpscaler(scale_method, model, vae, seed, steps, cfg, sampler_name, scheduler, + positive, negative, denoise, use_tiled_vae, upscale_model_opt, pk_hook_opt, + tile_size=tile_size, scheduler_func=scheduler_func_opt, + tile_cnet_opt=tile_cnet_opt, tile_cnet_strength=tile_cnet_strength) + return (upscaler, ) + + +class TwoSamplersForMaskUpscalerProvider: + upscale_methods = ["nearest-exact", "bilinear", "lanczos", "area"] + + @classmethod + def INPUT_TYPES(s): + return {"required": { + "scale_method": (s.upscale_methods,), + "full_sample_schedule": ( + ["none", "interleave1", "interleave2", "interleave3", + "last1", "last2", + "interleave1+last1", "interleave2+last1", "interleave3+last1", + ],), + "use_tiled_vae": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + "base_sampler": ("KSAMPLER", ), + "mask_sampler": ("KSAMPLER", ), + "mask": ("MASK", ), + "vae": ("VAE",), + "tile_size": ("INT", {"default": 512, "min": 320, "max": 4096, "step": 64}), + }, + "optional": { + "full_sampler_opt": ("KSAMPLER",), + "upscale_model_opt": ("UPSCALE_MODEL", ), + "pk_hook_base_opt": ("PK_HOOK", ), + "pk_hook_mask_opt": ("PK_HOOK", ), + "pk_hook_full_opt": ("PK_HOOK", ), + } + } + + RETURN_TYPES = ("UPSCALER", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Upscale" + + def doit(self, scale_method, full_sample_schedule, use_tiled_vae, base_sampler, mask_sampler, mask, vae, + full_sampler_opt=None, upscale_model_opt=None, + pk_hook_base_opt=None, pk_hook_mask_opt=None, pk_hook_full_opt=None, tile_size=512): + upscaler = core.TwoSamplersForMaskUpscaler(scale_method, full_sample_schedule, use_tiled_vae, + base_sampler, mask_sampler, mask, vae, full_sampler_opt, upscale_model_opt, + pk_hook_base_opt, pk_hook_mask_opt, pk_hook_full_opt, tile_size=tile_size) + return (upscaler, ) + + +class TwoSamplersForMaskUpscalerProviderPipe: + upscale_methods = ["nearest-exact", "bilinear", "lanczos", "area"] + + @classmethod + def INPUT_TYPES(s): + return {"required": { + "scale_method": (s.upscale_methods,), + "full_sample_schedule": ( + ["none", "interleave1", "interleave2", "interleave3", + "last1", "last2", + "interleave1+last1", "interleave2+last1", "interleave3+last1", + ],), + "use_tiled_vae": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + "base_sampler": ("KSAMPLER", ), + "mask_sampler": ("KSAMPLER", ), + "mask": ("MASK", ), + "basic_pipe": ("BASIC_PIPE",), + "tile_size": ("INT", {"default": 512, "min": 320, "max": 4096, "step": 64}), + }, + "optional": { + "full_sampler_opt": ("KSAMPLER",), + "upscale_model_opt": ("UPSCALE_MODEL", ), + "pk_hook_base_opt": ("PK_HOOK", ), + "pk_hook_mask_opt": ("PK_HOOK", ), + "pk_hook_full_opt": ("PK_HOOK", ), + } + } + + RETURN_TYPES = ("UPSCALER", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Upscale" + + def doit(self, scale_method, full_sample_schedule, use_tiled_vae, base_sampler, mask_sampler, mask, basic_pipe, + full_sampler_opt=None, upscale_model_opt=None, + pk_hook_base_opt=None, pk_hook_mask_opt=None, pk_hook_full_opt=None, tile_size=512): + + mask = make_2d_mask(mask) + + _, _, vae, _, _ = basic_pipe + upscaler = core.TwoSamplersForMaskUpscaler(scale_method, full_sample_schedule, use_tiled_vae, + base_sampler, mask_sampler, mask, vae, full_sampler_opt, upscale_model_opt, + pk_hook_base_opt, pk_hook_mask_opt, pk_hook_full_opt, tile_size=tile_size) + return (upscaler, ) + + +class IterativeLatentUpscale: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "samples": ("LATENT", ), + "upscale_factor": ("FLOAT", {"default": 1.5, "min": 1, "max": 10000, "step": 0.1}), + "steps": ("INT", {"default": 3, "min": 1, "max": 10000, "step": 1}), + "temp_prefix": ("STRING", {"default": ""}), + "upscaler": ("UPSCALER",), + "step_mode": (["simple", "geometric"], {"default": "simple"}) + }, + "hidden": {"unique_id": "UNIQUE_ID"}, + } + + RETURN_TYPES = ("LATENT", "VAE") + RETURN_NAMES = ("latent", "vae") + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Upscale" + + def doit(self, samples, upscale_factor, steps, temp_prefix, upscaler, step_mode="simple", unique_id=None): + w = samples['samples'].shape[3]*8 # image width + h = samples['samples'].shape[2]*8 # image height + + if temp_prefix == "": + temp_prefix = None + + if step_mode == "geometric": + upscale_factor_unit = pow(upscale_factor, 1.0/steps) + else: # simple + upscale_factor_unit = max(0, (upscale_factor - 1.0) / steps) + + current_latent = samples + noise_mask = current_latent.get('noise_mask') + scale = 1 + + for i in range(steps-1): + if step_mode == "geometric": + scale *= upscale_factor_unit + else: # simple + scale += upscale_factor_unit + + new_w = w*scale + new_h = h*scale + core.update_node_status(unique_id, f"{i+1}/{steps} steps | x{scale:.2f}", (i+1)/steps) + print(f"IterativeLatentUpscale[{i+1}/{steps}]: {new_w:.1f}x{new_h:.1f} (scale:{scale:.2f}) ") + step_info = i, steps + current_latent = upscaler.upscale_shape(step_info, current_latent, new_w, new_h, temp_prefix) + if noise_mask is not None: + current_latent['noise_mask'] = noise_mask + + if scale < upscale_factor: + new_w = w*upscale_factor + new_h = h*upscale_factor + core.update_node_status(unique_id, f"Final step | x{upscale_factor:.2f}", 1.0) + print(f"IterativeLatentUpscale[Final]: {new_w:.1f}x{new_h:.1f} (scale:{upscale_factor:.2f}) ") + step_info = steps-1, steps + current_latent = upscaler.upscale_shape(step_info, current_latent, new_w, new_h, temp_prefix) + + core.update_node_status(unique_id, "", None) + + return current_latent, upscaler.vae + + +class IterativeImageUpscale: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "pixels": ("IMAGE", ), + "upscale_factor": ("FLOAT", {"default": 1.5, "min": 1, "max": 10000, "step": 0.1}), + "steps": ("INT", {"default": 3, "min": 1, "max": 10000, "step": 1}), + "temp_prefix": ("STRING", {"default": ""}), + "upscaler": ("UPSCALER",), + "vae": ("VAE",), + "step_mode": (["simple", "geometric"], {"default": "simple"}) + }, + "hidden": {"unique_id": "UNIQUE_ID"} + } + + RETURN_TYPES = ("IMAGE",) + RETURN_NAMES = ("image",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Upscale" + + def doit(self, pixels, upscale_factor, steps, temp_prefix, upscaler, vae, step_mode="simple", unique_id=None): + if temp_prefix == "": + temp_prefix = None + + core.update_node_status(unique_id, "VAEEncode (first)", 0) + if upscaler.is_tiled: + latent = nodes.VAEEncodeTiled().encode(vae, pixels, upscaler.tile_size)[0] + else: + latent = nodes.VAEEncode().encode(vae, pixels)[0] + + refined_latent = IterativeLatentUpscale().doit(latent, upscale_factor, steps, temp_prefix, upscaler, step_mode, unique_id) + + core.update_node_status(unique_id, "VAEDecode (final)", 1.0) + if upscaler.is_tiled: + pixels = nodes.VAEDecodeTiled().decode(vae, refined_latent[0], upscaler.tile_size)[0] + else: + pixels = nodes.VAEDecode().decode(vae, refined_latent[0])[0] + + core.update_node_status(unique_id, "", None) + + return (pixels, ) + + +class FaceDetailerPipe: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "image": ("IMAGE", ), + "detailer_pipe": ("DETAILER_PIPE",), + "guide_size": ("FLOAT", {"default": 512, "min": 64, "max": nodes.MAX_RESOLUTION, "step": 8}), + "guide_size_for": ("BOOLEAN", {"default": True, "label_on": "bbox", "label_off": "crop_region"}), + "max_size": ("FLOAT", {"default": 1024, "min": 64, "max": nodes.MAX_RESOLUTION, "step": 8}), + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000}), + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS,), + "scheduler": (core.SCHEDULERS,), + "denoise": ("FLOAT", {"default": 0.5, "min": 0.0001, "max": 1.0, "step": 0.01}), + "feather": ("INT", {"default": 5, "min": 0, "max": 100, "step": 1}), + "noise_mask": ("BOOLEAN", {"default": True, "label_on": "enabled", "label_off": "disabled"}), + "force_inpaint": ("BOOLEAN", {"default": True, "label_on": "enabled", "label_off": "disabled"}), + + "bbox_threshold": ("FLOAT", {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01}), + "bbox_dilation": ("INT", {"default": 10, "min": -512, "max": 512, "step": 1}), + "bbox_crop_factor": ("FLOAT", {"default": 3.0, "min": 1.0, "max": 10, "step": 0.1}), + + "sam_detection_hint": (["center-1", "horizontal-2", "vertical-2", "rect-4", "diamond-4", "mask-area", "mask-points", "mask-point-bbox", "none"],), + "sam_dilation": ("INT", {"default": 0, "min": -512, "max": 512, "step": 1}), + "sam_threshold": ("FLOAT", {"default": 0.93, "min": 0.0, "max": 1.0, "step": 0.01}), + "sam_bbox_expansion": ("INT", {"default": 0, "min": 0, "max": 1000, "step": 1}), + "sam_mask_hint_threshold": ("FLOAT", {"default": 0.7, "min": 0.0, "max": 1.0, "step": 0.01}), + "sam_mask_hint_use_negative": (["False", "Small", "Outter"],), + + "drop_size": ("INT", {"min": 1, "max": MAX_RESOLUTION, "step": 1, "default": 10}), + "refiner_ratio": ("FLOAT", {"default": 0.2, "min": 0.0, "max": 1.0}), + + "cycle": ("INT", {"default": 1, "min": 1, "max": 10, "step": 1}), + }, + "optional": { + "inpaint_model": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + "noise_mask_feather": ("INT", {"default": 20, "min": 0, "max": 100, "step": 1}), + "scheduler_func_opt": ("SCHEDULER_FUNC",), + } + } + + RETURN_TYPES = ("IMAGE", "IMAGE", "IMAGE", "MASK", "DETAILER_PIPE", "IMAGE") + RETURN_NAMES = ("image", "cropped_refined", "cropped_enhanced_alpha", "mask", "detailer_pipe", "cnet_images") + OUTPUT_IS_LIST = (False, True, True, False, False, True) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Simple" + + def doit(self, image, detailer_pipe, guide_size, guide_size_for, max_size, seed, steps, cfg, sampler_name, scheduler, + denoise, feather, noise_mask, force_inpaint, bbox_threshold, bbox_dilation, bbox_crop_factor, + sam_detection_hint, sam_dilation, sam_threshold, sam_bbox_expansion, + sam_mask_hint_threshold, sam_mask_hint_use_negative, drop_size, refiner_ratio=None, + cycle=1, inpaint_model=False, noise_mask_feather=0, scheduler_func_opt=None): + + result_img = None + result_mask = None + result_cropped_enhanced = [] + result_cropped_enhanced_alpha = [] + result_cnet_images = [] + + if len(image) > 1: + print(f"[Impact Pack] WARN: FaceDetailer is not a node designed for video detailing. If you intend to perform video detailing, please use Detailer For AnimateDiff.") + + model, clip, vae, positive, negative, wildcard, bbox_detector, segm_detector, sam_model_opt, detailer_hook, \ + refiner_model, refiner_clip, refiner_positive, refiner_negative = detailer_pipe + + for i, single_image in enumerate(image): + enhanced_img, cropped_enhanced, cropped_enhanced_alpha, mask, cnet_pil_list = FaceDetailer.enhance_face( + single_image.unsqueeze(0), model, clip, vae, guide_size, guide_size_for, max_size, seed + i, steps, cfg, sampler_name, scheduler, + positive, negative, denoise, feather, noise_mask, force_inpaint, + bbox_threshold, bbox_dilation, bbox_crop_factor, + sam_detection_hint, sam_dilation, sam_threshold, sam_bbox_expansion, sam_mask_hint_threshold, + sam_mask_hint_use_negative, drop_size, bbox_detector, segm_detector, sam_model_opt, wildcard, detailer_hook, + refiner_ratio=refiner_ratio, refiner_model=refiner_model, + refiner_clip=refiner_clip, refiner_positive=refiner_positive, refiner_negative=refiner_negative, + cycle=cycle, inpaint_model=inpaint_model, noise_mask_feather=noise_mask_feather, scheduler_func_opt=scheduler_func_opt) + + result_img = torch.cat((result_img, enhanced_img), dim=0) if result_img is not None else enhanced_img + result_mask = torch.cat((result_mask, mask), dim=0) if result_mask is not None else mask + result_cropped_enhanced.extend(cropped_enhanced) + result_cropped_enhanced_alpha.extend(cropped_enhanced_alpha) + result_cnet_images.extend(cnet_pil_list) + + if len(result_cropped_enhanced) == 0: + result_cropped_enhanced = [empty_pil_tensor()] + + if len(result_cropped_enhanced_alpha) == 0: + result_cropped_enhanced_alpha = [empty_pil_tensor()] + + if len(result_cnet_images) == 0: + result_cnet_images = [empty_pil_tensor()] + + return result_img, result_cropped_enhanced, result_cropped_enhanced_alpha, result_mask, detailer_pipe, result_cnet_images + + +class MaskDetailerPipe: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "image": ("IMAGE", ), + "mask": ("MASK", ), + "basic_pipe": ("BASIC_PIPE",), + + "guide_size": ("FLOAT", {"default": 512, "min": 64, "max": nodes.MAX_RESOLUTION, "step": 8}), + "guide_size_for": ("BOOLEAN", {"default": True, "label_on": "mask bbox", "label_off": "crop region"}), + "max_size": ("FLOAT", {"default": 1024, "min": 64, "max": nodes.MAX_RESOLUTION, "step": 8}), + "mask_mode": ("BOOLEAN", {"default": True, "label_on": "masked only", "label_off": "whole"}), + + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000}), + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS,), + "scheduler": (core.SCHEDULERS,), + "denoise": ("FLOAT", {"default": 0.5, "min": 0.0001, "max": 1.0, "step": 0.01}), + + "feather": ("INT", {"default": 5, "min": 0, "max": 100, "step": 1}), + "crop_factor": ("FLOAT", {"default": 3.0, "min": 1.0, "max": 10, "step": 0.1}), + "drop_size": ("INT", {"min": 1, "max": MAX_RESOLUTION, "step": 1, "default": 10}), + "refiner_ratio": ("FLOAT", {"default": 0.2, "min": 0.0, "max": 1.0}), + "batch_size": ("INT", {"default": 1, "min": 1, "max": 100}), + + "cycle": ("INT", {"default": 1, "min": 1, "max": 10, "step": 1}), + }, + "optional": { + "refiner_basic_pipe_opt": ("BASIC_PIPE", ), + "detailer_hook": ("DETAILER_HOOK",), + "inpaint_model": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + "noise_mask_feather": ("INT", {"default": 20, "min": 0, "max": 100, "step": 1}), + "bbox_fill": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + "contour_fill": ("BOOLEAN", {"default": True, "label_on": "enabled", "label_off": "disabled"}), + "scheduler_func_opt": ("SCHEDULER_FUNC",), + } + } + + RETURN_TYPES = ("IMAGE", "IMAGE", "IMAGE", "BASIC_PIPE", "BASIC_PIPE") + RETURN_NAMES = ("image", "cropped_refined", "cropped_enhanced_alpha", "basic_pipe", "refiner_basic_pipe_opt") + OUTPUT_IS_LIST = (False, True, True, False, False) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detailer" + + def doit(self, image, mask, basic_pipe, guide_size, guide_size_for, max_size, mask_mode, + seed, steps, cfg, sampler_name, scheduler, denoise, + feather, crop_factor, drop_size, refiner_ratio, batch_size, cycle=1, + refiner_basic_pipe_opt=None, detailer_hook=None, inpaint_model=False, noise_mask_feather=0, + bbox_fill=False, contour_fill=True, scheduler_func_opt=None): + + if len(image) > 1: + raise Exception('[Impact Pack] ERROR: MaskDetailer does not allow image batches.\nPlease refer to https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/batching-detailer.md for more information.') + + model, clip, vae, positive, negative = basic_pipe + + if refiner_basic_pipe_opt is None: + refiner_model, refiner_clip, refiner_positive, refiner_negative = None, None, None, None + else: + refiner_model, refiner_clip, _, refiner_positive, refiner_negative = refiner_basic_pipe_opt + + # create segs + if mask is not None: + mask = make_2d_mask(mask) + segs = core.mask_to_segs(mask, False, crop_factor, bbox_fill, drop_size, is_contour=contour_fill) + else: + segs = ((image.shape[1], image.shape[2]), []) + + enhanced_img_batch = None + cropped_enhanced_list = [] + cropped_enhanced_alpha_list = [] + + for i in range(batch_size): + if mask is not None: + enhanced_img, _, cropped_enhanced, cropped_enhanced_alpha, _, _ = \ + DetailerForEach.do_detail(image, segs, model, clip, vae, guide_size, guide_size_for, max_size, seed+i, steps, + cfg, sampler_name, scheduler, positive, negative, denoise, feather, mask_mode, + force_inpaint=True, wildcard_opt=None, detailer_hook=detailer_hook, + refiner_ratio=refiner_ratio, refiner_model=refiner_model, refiner_clip=refiner_clip, + refiner_positive=refiner_positive, refiner_negative=refiner_negative, + cycle=cycle, inpaint_model=inpaint_model, noise_mask_feather=noise_mask_feather, scheduler_func_opt=scheduler_func_opt) + else: + enhanced_img, cropped_enhanced, cropped_enhanced_alpha = image, [], [] + + if enhanced_img_batch is None: + enhanced_img_batch = enhanced_img + else: + enhanced_img_batch = torch.cat((enhanced_img_batch, enhanced_img), dim=0) + + cropped_enhanced_list += cropped_enhanced + cropped_enhanced_alpha_list += cropped_enhanced_alpha + + # set fallback image + if len(cropped_enhanced_list) == 0: + cropped_enhanced_list = [empty_pil_tensor()] + + if len(cropped_enhanced_alpha_list) == 0: + cropped_enhanced_alpha_list = [empty_pil_tensor()] + + return enhanced_img_batch, cropped_enhanced_list, cropped_enhanced_alpha_list, basic_pipe, refiner_basic_pipe_opt + + +class DetailerForEachTest(DetailerForEach): + RETURN_TYPES = ("IMAGE", "IMAGE", "IMAGE", "IMAGE", "IMAGE") + RETURN_NAMES = ("image", "cropped", "cropped_refined", "cropped_refined_alpha", "cnet_images") + OUTPUT_IS_LIST = (False, True, True, True, True) + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detailer" + + def doit(self, image, segs, model, clip, vae, guide_size, guide_size_for, max_size, seed, steps, cfg, sampler_name, + scheduler, positive, negative, denoise, feather, noise_mask, force_inpaint, wildcard, detailer_hook=None, + cycle=1, inpaint_model=False, noise_mask_feather=0, scheduler_func_opt=None): + + if len(image) > 1: + raise Exception('[Impact Pack] ERROR: DetailerForEach does not allow image batches.\nPlease refer to https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/batching-detailer.md for more information.') + + enhanced_img, cropped, cropped_enhanced, cropped_enhanced_alpha, cnet_pil_list, new_segs = \ + DetailerForEach.do_detail(image, segs, model, clip, vae, guide_size, guide_size_for, max_size, seed, steps, + cfg, sampler_name, scheduler, positive, negative, denoise, feather, noise_mask, + force_inpaint, wildcard, detailer_hook, + cycle=cycle, inpaint_model=inpaint_model, noise_mask_feather=noise_mask_feather, scheduler_func_opt=scheduler_func_opt) + + # set fallback image + if len(cropped) == 0: + cropped = [empty_pil_tensor()] + + if len(cropped_enhanced) == 0: + cropped_enhanced = [empty_pil_tensor()] + + if len(cropped_enhanced_alpha) == 0: + cropped_enhanced_alpha = [empty_pil_tensor()] + + if len(cnet_pil_list) == 0: + cnet_pil_list = [empty_pil_tensor()] + + return enhanced_img, cropped, cropped_enhanced, cropped_enhanced_alpha, cnet_pil_list + + +class DetailerForEachTestPipe(DetailerForEachPipe): + RETURN_TYPES = ("IMAGE", "SEGS", "BASIC_PIPE", "IMAGE", "IMAGE", "IMAGE", "IMAGE", ) + RETURN_NAMES = ("image", "segs", "basic_pipe", "cropped", "cropped_refined", "cropped_refined_alpha", 'cnet_images') + OUTPUT_IS_LIST = (False, False, False, True, True, True, True) + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detailer" + + def doit(self, image, segs, guide_size, guide_size_for, max_size, seed, steps, cfg, sampler_name, scheduler, + denoise, feather, noise_mask, force_inpaint, basic_pipe, wildcard, cycle=1, + refiner_ratio=None, detailer_hook=None, refiner_basic_pipe_opt=None, inpaint_model=False, noise_mask_feather=0, scheduler_func_opt=None): + + if len(image) > 1: + raise Exception('[Impact Pack] ERROR: DetailerForEach does not allow image batches.\nPlease refer to https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/batching-detailer.md for more information.') + + model, clip, vae, positive, negative = basic_pipe + + if refiner_basic_pipe_opt is None: + refiner_model, refiner_clip, refiner_positive, refiner_negative = None, None, None, None + else: + refiner_model, refiner_clip, _, refiner_positive, refiner_negative = refiner_basic_pipe_opt + + enhanced_img, cropped, cropped_enhanced, cropped_enhanced_alpha, cnet_pil_list, new_segs = \ + DetailerForEach.do_detail(image, segs, model, clip, vae, guide_size, guide_size_for, max_size, seed, steps, cfg, + sampler_name, scheduler, positive, negative, denoise, feather, noise_mask, + force_inpaint, wildcard, detailer_hook, + refiner_ratio=refiner_ratio, refiner_model=refiner_model, + refiner_clip=refiner_clip, refiner_positive=refiner_positive, + refiner_negative=refiner_negative, + cycle=cycle, inpaint_model=inpaint_model, noise_mask_feather=noise_mask_feather, scheduler_func_opt=scheduler_func_opt) + + # set fallback image + if len(cropped) == 0: + cropped = [empty_pil_tensor()] + + if len(cropped_enhanced) == 0: + cropped_enhanced = [empty_pil_tensor()] + + if len(cropped_enhanced_alpha) == 0: + cropped_enhanced_alpha = [empty_pil_tensor()] + + if len(cnet_pil_list) == 0: + cnet_pil_list = [empty_pil_tensor()] + + return enhanced_img, new_segs, basic_pipe, cropped, cropped_enhanced, cropped_enhanced_alpha, cnet_pil_list + + +class SegsBitwiseAndMask: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS",), + "mask": ("MASK",), + } + } + + RETURN_TYPES = ("SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Operation" + + def doit(self, segs, mask): + return (core.segs_bitwise_and_mask(segs, mask), ) + + +class SegsBitwiseAndMaskForEach: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS",), + "masks": ("MASK",), + } + } + + RETURN_TYPES = ("SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Operation" + + def doit(self, segs, masks): + return (core.apply_mask_to_each_seg(segs, masks), ) + + +class BitwiseAndMaskForEach: + @classmethod + def INPUT_TYPES(s): + return {"required": + { + "base_segs": ("SEGS",), + "mask_segs": ("SEGS",), + } + } + + RETURN_TYPES = ("SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Operation" + + DESCRIPTION = "Retains only the overlapping areas between the masks included in base_segs and the mask regions of mask_segs. SEGS with no overlapping mask areas are filtered out." + + def doit(self, base_segs, mask_segs): + mask = core.segs_to_combined_mask(mask_segs) + mask = make_3d_mask(mask) + + return SegsBitwiseAndMask().doit(base_segs, mask) + + +class SubtractMaskForEach: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "base_segs": ("SEGS",), + "mask_segs": ("SEGS",), + } + } + + RETURN_TYPES = ("SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Operation" + + DESCRIPTION = "Removes only the overlapping areas between the masks included in base_segs and the mask regions of mask_segs. SEGS with no overlapping mask areas are filtered out." + + def doit(self, base_segs, mask_segs): + mask = core.segs_to_combined_mask(mask_segs) + mask = make_3d_mask(mask) + return (core.segs_bitwise_subtract_mask(base_segs, mask), ) + + +class ToBinaryMask: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "mask": ("MASK",), + "threshold": ("INT", {"default": 20, "min": 1, "max": 255}), + } + } + + RETURN_TYPES = ("MASK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Operation" + + def doit(self, mask, threshold): + mask = to_binary_mask(mask, threshold/255.0) + return (mask,) + + +class FlattenMask: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "masks": ("MASK",), + } + } + + RETURN_TYPES = ("MASK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Operation" + + def doit(self, masks): + masks = utils.make_3d_mask(masks) + masks = utils.flatten_mask(masks) + return (masks,) + + +class BitwiseAndMask: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "mask1": ("MASK",), + "mask2": ("MASK",), + } + } + + RETURN_TYPES = ("MASK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Operation" + + def doit(self, mask1, mask2): + mask = bitwise_and_masks(mask1, mask2) + return (mask,) + + +class SubtractMask: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "mask1": ("MASK", ), + "mask2": ("MASK", ), + } + } + + RETURN_TYPES = ("MASK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Operation" + + def doit(self, mask1, mask2): + mask = subtract_masks(mask1, mask2) + return (mask,) + + +class AddMask: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "mask1": ("MASK",), + "mask2": ("MASK",), + } + } + + RETURN_TYPES = ("MASK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Operation" + + def doit(self, mask1, mask2): + mask = add_masks(mask1, mask2) + return (mask,) + + +import nodes + + +def get_image_hash(arr): + split_index1 = arr.shape[0] // 2 + split_index2 = arr.shape[1] // 2 + part1 = arr[:split_index1, :split_index2] + part2 = arr[:split_index1, split_index2:] + part3 = arr[split_index1:, :split_index2] + part4 = arr[split_index1:, split_index2:] + + # 각 부분을 합산 + sum1 = np.sum(part1) + sum2 = np.sum(part2) + sum3 = np.sum(part3) + sum4 = np.sum(part4) + + return hash((sum1, sum2, sum3, sum4)) + + +def get_file_item(base_type, path): + path_type = base_type + + if path == "[output]": + path_type = "output" + path = path[:-9] + elif path == "[input]": + path_type = "input" + path = path[:-8] + elif path == "[temp]": + path_type = "temp" + path = path[:-7] + + subfolder = os.path.dirname(path) + filename = os.path.basename(path) + + return { + "filename": filename, + "subfolder": subfolder, + "type": path_type + } + + +class ImageReceiver: + @classmethod + def INPUT_TYPES(s): + input_dir = folder_paths.get_input_directory() + files = [f for f in os.listdir(input_dir) if os.path.isfile(os.path.join(input_dir, f))] + return {"required": { + "image": (sorted(files), ), + "link_id": ("INT", {"default": 0, "min": 0, "max": sys.maxsize, "step": 1}), + "save_to_workflow": ("BOOLEAN", {"default": False}), + "image_data": ("STRING", {"multiline": False}), + "trigger_always": ("BOOLEAN", {"default": False, "label_on": "enable", "label_off": "disable"}), + }, + } + + FUNCTION = "doit" + + RETURN_TYPES = ("IMAGE", "MASK") + + CATEGORY = "ImpactPack/Util" + + def doit(self, image, link_id, save_to_workflow, image_data, trigger_always): + if save_to_workflow: + try: + image_data = base64.b64decode(image_data.split(",")[1]) + i = Image.open(BytesIO(image_data)) + i = ImageOps.exif_transpose(i) + image = i.convert("RGB") + image = np.array(image).astype(np.float32) / 255.0 + image = torch.from_numpy(image)[None,] + if 'A' in i.getbands(): + mask = np.array(i.getchannel('A')).astype(np.float32) / 255.0 + mask = 1. - torch.from_numpy(mask) + else: + mask = torch.zeros((64, 64), dtype=torch.float32, device="cpu") + return (image, mask.unsqueeze(0)) + except Exception as e: + print(f"[WARN] ComfyUI-Impact-Pack: ImageReceiver - invalid 'image_data'") + mask = torch.zeros((64, 64), dtype=torch.float32, device="cpu") + return (empty_pil_tensor(64, 64), mask, ) + else: + return nodes.LoadImage().load_image(image) + + @classmethod + def VALIDATE_INPUTS(s, image, link_id, save_to_workflow, image_data, trigger_always): + if image != '#DATA' and not folder_paths.exists_annotated_filepath(image) or image.startswith("/") or ".." in image: + return "Invalid image file: {}".format(image) + + return True + + @classmethod + def IS_CHANGED(s, image, link_id, save_to_workflow, image_data, trigger_always): + if trigger_always: + return float("NaN") + else: + if save_to_workflow: + return hash(image_data) + else: + return hash(image) + + +from server import PromptServer + +class ImageSender(nodes.PreviewImage): + @classmethod + def INPUT_TYPES(s): + return {"required": { + "images": ("IMAGE", ), + "filename_prefix": ("STRING", {"default": "ImgSender"}), + "link_id": ("INT", {"default": 0, "min": 0, "max": sys.maxsize, "step": 1}), }, + "hidden": {"prompt": "PROMPT", "extra_pnginfo": "EXTRA_PNGINFO"}, + } + + OUTPUT_NODE = True + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, images, filename_prefix="ImgSender", link_id=0, prompt=None, extra_pnginfo=None): + result = nodes.PreviewImage().save_images(images, filename_prefix, prompt, extra_pnginfo) + PromptServer.instance.send_sync("img-send", {"link_id": link_id, "images": result['ui']['images']}) + return result + + +class LatentReceiver: + def __init__(self): + self.input_dir = folder_paths.get_input_directory() + self.type = "input" + + @classmethod + def INPUT_TYPES(s): + def check_file_extension(x): + return x.endswith(".latent") or x.endswith(".latent.png") + + input_dir = folder_paths.get_input_directory() + files = [f for f in os.listdir(input_dir) if os.path.isfile(os.path.join(input_dir, f)) and check_file_extension(f)] + return {"required": { + "latent": (sorted(files), ), + "link_id": ("INT", {"default": 0, "min": 0, "max": sys.maxsize, "step": 1}), + "trigger_always": ("BOOLEAN", {"default": False, "label_on": "enable", "label_off": "disable"}), + }, + } + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + RETURN_TYPES = ("LATENT",) + + @staticmethod + def load_preview_latent(image_path): + if not os.path.exists(image_path): + return None + + image = Image.open(image_path) + exif_data = piexif.load(image.info["exif"]) + + if piexif.ExifIFD.UserComment in exif_data["Exif"]: + compressed_data = exif_data["Exif"][piexif.ExifIFD.UserComment] + compressed_data_io = BytesIO(compressed_data) + with zipfile.ZipFile(compressed_data_io, mode='r') as archive: + tensor_bytes = archive.read("latent") + tensor = safetensors.torch.load(tensor_bytes) + return {"samples": tensor['latent_tensor']} + return None + + def parse_filename(self, filename): + pattern = r"^(.*)/(.*?)\[(.*)\]\s*$" + match = re.match(pattern, filename) + if match: + subfolder = match.group(1) + filename = match.group(2).rstrip() + file_type = match.group(3) + else: + subfolder = '' + file_type = self.type + + return {'filename': filename, 'subfolder': subfolder, 'type': file_type} + + def doit(self, **kwargs): + if 'latent' not in kwargs: + return (torch.zeros([1, 4, 8, 8]), ) + + latent = kwargs['latent'] + + latent_name = latent + latent_path = folder_paths.get_annotated_filepath(latent_name) + + if latent.endswith(".latent"): + latent = safetensors.torch.load_file(latent_path, device="cpu") + multiplier = 1.0 + if "latent_format_version_0" not in latent: + multiplier = 1.0 / 0.18215 + samples = {"samples": latent["latent_tensor"].float() * multiplier} + else: + samples = LatentReceiver.load_preview_latent(latent_path) + + if samples is None: + samples = {'samples': torch.zeros([1, 4, 8, 8])} + + preview = self.parse_filename(latent_name) + + return { + 'ui': {"images": [preview]}, + 'result': (samples, ) + } + + @classmethod + def IS_CHANGED(s, latent, link_id, trigger_always): + if trigger_always: + return float("NaN") + else: + image_path = folder_paths.get_annotated_filepath(latent) + m = hashlib.sha256() + with open(image_path, 'rb') as f: + m.update(f.read()) + return m.digest().hex() + + @classmethod + def VALIDATE_INPUTS(s, latent, link_id, trigger_always): + if not folder_paths.exists_annotated_filepath(latent) or latent.startswith("/") or ".." in latent: + return "Invalid latent file: {}".format(latent) + return True + + +class LatentSender(nodes.SaveLatent): + def __init__(self): + super().__init__() + self.output_dir = folder_paths.get_temp_directory() + self.type = "temp" + + @classmethod + def INPUT_TYPES(s): + return {"required": { + "samples": ("LATENT", ), + "filename_prefix": ("STRING", {"default": "latents/LatentSender"}), + "link_id": ("INT", {"default": 0, "min": 0, "max": sys.maxsize, "step": 1}), + "preview_method": (["Latent2RGB-SDXL", "Latent2RGB-SD15", "TAESDXL", "TAESD15"],) + }, + "hidden": {"prompt": "PROMPT", "extra_pnginfo": "EXTRA_PNGINFO"}, + } + + OUTPUT_NODE = True + + RETURN_TYPES = () + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + @staticmethod + def save_to_file(tensor_bytes, prompt, extra_pnginfo, image, image_path): + compressed_data = BytesIO() + with zipfile.ZipFile(compressed_data, mode='w') as archive: + archive.writestr("latent", tensor_bytes) + image = image.copy() + exif_data = {"Exif": {piexif.ExifIFD.UserComment: compressed_data.getvalue()}} + + metadata = PngInfo() + if prompt is not None: + metadata.add_text("prompt", json.dumps(prompt)) + if extra_pnginfo is not None: + for x in extra_pnginfo: + metadata.add_text(x, json.dumps(extra_pnginfo[x])) + + exif_bytes = piexif.dump(exif_data) + image.save(image_path, format='png', exif=exif_bytes, pnginfo=metadata, optimize=True) + + @staticmethod + def prepare_preview(latent_tensor, preview_method): + from comfy.cli_args import LatentPreviewMethod + import comfy.latent_formats as latent_formats + + lower_bound = 128 + upper_bound = 256 + + if preview_method == "Latent2RGB-SD15": + latent_format = latent_formats.SD15() + method = LatentPreviewMethod.Latent2RGB + elif preview_method == "TAESD15": + latent_format = latent_formats.SD15() + method = LatentPreviewMethod.TAESD + elif preview_method == "TAESDXL": + latent_format = latent_formats.SDXL() + method = LatentPreviewMethod.TAESD + else: # preview_method == "Latent2RGB-SDXL" + latent_format = latent_formats.SDXL() + method = LatentPreviewMethod.Latent2RGB + + previewer = core.get_previewer("cpu", latent_format=latent_format, force=True, method=method) + + image = previewer.decode_latent_to_preview(latent_tensor) + min_size = min(image.size[0], image.size[1]) + max_size = max(image.size[0], image.size[1]) + + scale_factor = 1 + if max_size > upper_bound: + scale_factor = upper_bound/max_size + + # prevent too small preview + if min_size*scale_factor < lower_bound: + scale_factor = lower_bound/min_size + + w = int(image.size[0] * scale_factor) + h = int(image.size[1] * scale_factor) + + image = image.resize((w, h), resample=Image.NEAREST) + + return LatentSender.attach_format_text(image) + + @staticmethod + def attach_format_text(image): + width_a, height_a = image.size + + letter_image = Image.open(latent_letter_path) + width_b, height_b = letter_image.size + + new_width = max(width_a, width_b) + new_height = height_a + height_b + + new_image = Image.new('RGB', (new_width, new_height), (0, 0, 0)) + + offset_x = (new_width - width_b) // 2 + offset_y = (height_a + (new_height - height_a - height_b) // 2) + new_image.paste(letter_image, (offset_x, offset_y)) + + new_image.paste(image, (0, 0)) + + return new_image + + def doit(self, samples, filename_prefix="latents/LatentSender", link_id=0, preview_method="Latent2RGB-SDXL", prompt=None, extra_pnginfo=None): + full_output_folder, filename, counter, subfolder, filename_prefix = folder_paths.get_save_image_path(filename_prefix, self.output_dir) + + # load preview + preview = LatentSender.prepare_preview(samples['samples'], preview_method) + + # support save metadata for latent sharing + file = f"{filename}_{counter:05}_.latent.png" + fullpath = os.path.join(full_output_folder, file) + + output = {"latent_tensor": samples["samples"]} + + tensor_bytes = safetensors.torch.save(output) + LatentSender.save_to_file(tensor_bytes, prompt, extra_pnginfo, preview, fullpath) + + latent_path = { + 'filename': file, + 'subfolder': subfolder, + 'type': self.type + } + + PromptServer.instance.send_sync("latent-send", {"link_id": link_id, "images": [latent_path]}) + + return {'ui': {'images': [latent_path]}} + + +class ImpactWildcardProcessor: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "wildcard_text": ("STRING", {"multiline": True, "dynamicPrompts": False}), + "populated_text": ("STRING", {"multiline": True, "dynamicPrompts": False}), + "mode": ("BOOLEAN", {"default": True, "label_on": "Populate", "label_off": "Fixed"}), + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "Select to add Wildcard": (["Select the Wildcard to add to the text"],), + }, + } + + CATEGORY = "ImpactPack/Prompt" + + RETURN_TYPES = ("STRING", ) + FUNCTION = "doit" + + @staticmethod + def process(**kwargs): + return impact.wildcards.process(**kwargs) + + def doit(self, *args, **kwargs): + populated_text = ImpactWildcardProcessor.process(text=kwargs['populated_text'], seed=kwargs['seed']) + return (populated_text, ) + + +class ImpactWildcardEncode: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "model": ("MODEL",), + "clip": ("CLIP",), + "wildcard_text": ("STRING", {"multiline": True, "dynamicPrompts": False}), + "populated_text": ("STRING", {"multiline": True, "dynamicPrompts": False}), + "mode": ("BOOLEAN", {"default": True, "label_on": "Populate", "label_off": "Fixed"}), + "Select to add LoRA": (["Select the LoRA to add to the text"] + folder_paths.get_filename_list("loras"), ), + "Select to add Wildcard": (["Select the Wildcard to add to the text"], ), + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + }, + } + + CATEGORY = "ImpactPack/Prompt" + + RETURN_TYPES = ("MODEL", "CLIP", "CONDITIONING", "STRING") + RETURN_NAMES = ("model", "clip", "conditioning", "populated_text") + FUNCTION = "doit" + + @staticmethod + def process_with_loras(**kwargs): + return impact.wildcards.process_with_loras(**kwargs) + + @staticmethod + def get_wildcard_list(): + return impact.wildcards.get_wildcard_list() + + def doit(self, *args, **kwargs): + populated = kwargs['populated_text'] + processed = [] + model, clip, conditioning = impact.wildcards.process_with_loras(wildcard_opt=populated, model=kwargs['model'], clip=kwargs['clip'], seed=kwargs['seed'], processed=processed) + return model, clip, conditioning, processed[0] + + +class ImpactSchedulerAdapter: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "scheduler": (comfy.samplers.KSampler.SCHEDULERS, {"defaultInput": True, }), + "extra_scheduler": (['None', 'AYS SDXL', 'AYS SD1', 'AYS SVD', 'GITS[coeff=1.2]'],), + }} + + CATEGORY = "ImpactPack/Util" + + RETURN_TYPES = (core.SCHEDULERS,) + RETURN_NAMES = ("scheduler",) + + FUNCTION = "doit" + + def doit(self, scheduler, extra_scheduler): + if extra_scheduler != 'None': + return (extra_scheduler,) + + return (scheduler,) + diff --git a/ComfyUI-Impact-Pack/modules/impact/impact_sampling.py b/ComfyUI-Impact-Pack/modules/impact/impact_sampling.py new file mode 100644 index 0000000000000000000000000000000000000000..4030d02b4c4f0fb303264c53d509511da3ff2ff4 --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/impact_sampling.py @@ -0,0 +1,355 @@ +import nodes +from comfy.k_diffusion import sampling as k_diffusion_sampling +from comfy import samplers +from comfy_extras import nodes_custom_sampler +import latent_preview +import comfy +import torch +import math +import comfy.model_management as mm + + +try: + from comfy_extras.nodes_custom_sampler import Noise_EmptyNoise, Noise_RandomNoise + import node_helpers +except: + print(f"\n#############################################\n[Impact Pack] ComfyUI is an outdated version.\n#############################################\n") + raise Exception("[Impact Pack] ComfyUI is an outdated version.") + + +def calculate_sigmas(model, sampler, scheduler, steps): + discard_penultimate_sigma = False + if sampler in ['dpm_2', 'dpm_2_ancestral', 'uni_pc', 'uni_pc_bh2']: + steps += 1 + discard_penultimate_sigma = True + + if scheduler.startswith('AYS'): + sigmas = nodes.NODE_CLASS_MAPPINGS['AlignYourStepsScheduler']().get_sigmas(scheduler[4:], steps, denoise=1.0)[0] + elif scheduler.startswith('GITS[coeff='): + sigmas = nodes.NODE_CLASS_MAPPINGS['GITSScheduler']().get_sigmas(float(scheduler[11:-1]), steps, denoise=1.0)[0] + else: + sigmas = samplers.calculate_sigmas(model.get_model_object("model_sampling"), scheduler, steps) + + if discard_penultimate_sigma: + sigmas = torch.cat([sigmas[:-2], sigmas[-1:]]) + return sigmas + + +def get_noise_sampler(x, cpu, total_sigmas, **kwargs): + if 'extra_args' in kwargs and 'seed' in kwargs['extra_args']: + sigma_min, sigma_max = total_sigmas[total_sigmas > 0].min(), total_sigmas.max() + seed = kwargs['extra_args'].get("seed", None) + return k_diffusion_sampling.BrownianTreeNoiseSampler(x, sigma_min, sigma_max, seed=seed, cpu=cpu) + return None + + +def ksampler(sampler_name, total_sigmas, extra_options={}, inpaint_options={}): + if sampler_name == "dpmpp_sde": + def sample_dpmpp_sde(model, x, sigmas, **kwargs): + noise_sampler = get_noise_sampler(x, True, total_sigmas, **kwargs) + if noise_sampler is not None: + kwargs['noise_sampler'] = noise_sampler + + return k_diffusion_sampling.sample_dpmpp_sde(model, x, sigmas, **kwargs) + + sampler_function = sample_dpmpp_sde + + elif sampler_name == "dpmpp_sde_gpu": + def sample_dpmpp_sde(model, x, sigmas, **kwargs): + noise_sampler = get_noise_sampler(x, False, total_sigmas, **kwargs) + if noise_sampler is not None: + kwargs['noise_sampler'] = noise_sampler + + return k_diffusion_sampling.sample_dpmpp_sde_gpu(model, x, sigmas, **kwargs) + + sampler_function = sample_dpmpp_sde + + elif sampler_name == "dpmpp_2m_sde": + def sample_dpmpp_sde(model, x, sigmas, **kwargs): + noise_sampler = get_noise_sampler(x, True, total_sigmas, **kwargs) + if noise_sampler is not None: + kwargs['noise_sampler'] = noise_sampler + + return k_diffusion_sampling.sample_dpmpp_2m_sde(model, x, sigmas, **kwargs) + + sampler_function = sample_dpmpp_sde + + elif sampler_name == "dpmpp_2m_sde_gpu": + def sample_dpmpp_sde(model, x, sigmas, **kwargs): + noise_sampler = get_noise_sampler(x, False, total_sigmas, **kwargs) + if noise_sampler is not None: + kwargs['noise_sampler'] = noise_sampler + + return k_diffusion_sampling.sample_dpmpp_2m_sde_gpu(model, x, sigmas, **kwargs) + + sampler_function = sample_dpmpp_sde + + elif sampler_name == "dpmpp_3m_sde": + def sample_dpmpp_sde(model, x, sigmas, **kwargs): + noise_sampler = get_noise_sampler(x, True, total_sigmas, **kwargs) + if noise_sampler is not None: + kwargs['noise_sampler'] = noise_sampler + + return k_diffusion_sampling.sample_dpmpp_3m_sde(model, x, sigmas, **kwargs) + + sampler_function = sample_dpmpp_sde + + elif sampler_name == "dpmpp_3m_sde_gpu": + def sample_dpmpp_sde(model, x, sigmas, **kwargs): + noise_sampler = get_noise_sampler(x, False, total_sigmas, **kwargs) + if noise_sampler is not None: + kwargs['noise_sampler'] = noise_sampler + + return k_diffusion_sampling.sample_dpmpp_3m_sde_gpu(model, x, sigmas, **kwargs) + + sampler_function = sample_dpmpp_sde + + else: + return comfy.samplers.sampler_object(sampler_name) + + return samplers.KSAMPLER(sampler_function, extra_options, inpaint_options) + + +# modified version of SamplerCustom.sample +def sample_with_custom_noise(model, add_noise, noise_seed, cfg, positive, negative, sampler, sigmas, latent_image, noise=None, callback=None): + latent = latent_image + latent_image = latent["samples"] + + if hasattr(comfy.sample, 'fix_empty_latent_channels'): + latent_image = comfy.sample.fix_empty_latent_channels(model, latent_image) + + out = latent.copy() + out['samples'] = latent_image + + if noise is None: + if not add_noise: + noise = Noise_EmptyNoise().generate_noise(out) + else: + noise = Noise_RandomNoise(noise_seed).generate_noise(out) + + noise_mask = None + if "noise_mask" in latent: + noise_mask = latent["noise_mask"] + + x0_output = {} + preview_callback = latent_preview.prepare_callback(model, sigmas.shape[-1] - 1, x0_output) + + if callback is not None: + def touched_callback(step, x0, x, total_steps): + callback(step, x0, x, total_steps) + preview_callback(step, x0, x, total_steps) + else: + touched_callback = preview_callback + + disable_pbar = not comfy.utils.PROGRESS_BAR_ENABLED + + device = mm.get_torch_device() + + noise = noise.to(device) + latent_image = latent_image.to(device) + if noise_mask is not None: + noise_mask = noise_mask.to(device) + + if negative != 'NegativePlaceholder': + # This way is incompatible with Advanced ControlNet, yet. + # guider = comfy.samplers.CFGGuider(model) + # guider.set_conds(positive, negative) + # guider.set_cfg(cfg) + samples = comfy.sample.sample_custom(model, noise, cfg, sampler, sigmas, positive, negative, latent_image, + noise_mask=noise_mask, callback=touched_callback, + disable_pbar=disable_pbar, seed=noise_seed) + else: + guider = nodes_custom_sampler.Guider_Basic(model) + positive = node_helpers.conditioning_set_values(positive, {"guidance": cfg}) + guider.set_conds(positive) + samples = guider.sample(noise, latent_image, sampler, sigmas, denoise_mask=noise_mask, callback=touched_callback, disable_pbar=disable_pbar, seed=noise_seed) + + samples = samples.to(comfy.model_management.intermediate_device()) + + out["samples"] = samples + if "x0" in x0_output: + out_denoised = latent.copy() + out_denoised["samples"] = model.model.process_latent_out(x0_output["x0"].cpu()) + else: + out_denoised = out + return out, out_denoised + + +# When sampling one step at a time, it mitigates the problem. (especially for _sde series samplers) +def separated_sample(model, add_noise, seed, steps, cfg, sampler_name, scheduler, positive, negative, + latent_image, start_at_step, end_at_step, return_with_leftover_noise, sigma_ratio=1.0, sampler_opt=None, noise=None, callback=None, scheduler_func=None): + + if scheduler_func is not None: + total_sigmas = scheduler_func(model, sampler_name, steps) + else: + if sampler_opt is None: + total_sigmas = calculate_sigmas(model, sampler_name, scheduler, steps) + else: + total_sigmas = calculate_sigmas(model, "", scheduler, steps) + + sigmas = total_sigmas + + if end_at_step is not None and end_at_step < (len(total_sigmas) - 1): + sigmas = total_sigmas[:end_at_step + 1] + if not return_with_leftover_noise: + sigmas[-1] = 0 + + if start_at_step is not None: + if start_at_step < (len(sigmas) - 1): + sigmas = sigmas[start_at_step:] * sigma_ratio + else: + if latent_image is not None: + return latent_image + else: + return {'samples': torch.zeros_like(noise)} + + if sampler_opt is None: + impact_sampler = ksampler(sampler_name, total_sigmas) + else: + impact_sampler = sampler_opt + + if len(sigmas) == 0 or (len(sigmas) == 1 and sigmas[0] == 0): + return latent_image + + res = sample_with_custom_noise(model, add_noise, seed, cfg, positive, negative, impact_sampler, sigmas, latent_image, noise=noise, callback=callback) + + if return_with_leftover_noise: + return res[0] + else: + return res[1] + + +def impact_sample(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise=1.0, sigma_ratio=1.0, sampler_opt=None, noise=None, scheduler_func=None): + advanced_steps = math.floor(steps / denoise) + start_at_step = advanced_steps - steps + end_at_step = start_at_step + steps + return separated_sample(model, True, seed, advanced_steps, cfg, sampler_name, scheduler, positive, negative, latent_image, + start_at_step, end_at_step, False, scheduler_func=scheduler_func) + + +def ksampler_wrapper(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise, + refiner_ratio=None, refiner_model=None, refiner_clip=None, refiner_positive=None, refiner_negative=None, sigma_factor=1.0, noise=None, scheduler_func=None): + + if refiner_ratio is None or refiner_model is None or refiner_clip is None or refiner_positive is None or refiner_negative is None: + # Use separated_sample instead of KSampler for `AYS scheduler` + # refined_latent = nodes.KSampler().sample(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise * sigma_factor)[0] + + advanced_steps = math.floor(steps / denoise) + start_at_step = advanced_steps - steps + end_at_step = start_at_step + steps + + refined_latent = separated_sample(model, True, seed, advanced_steps, cfg, sampler_name, scheduler, + positive, negative, latent_image, start_at_step, end_at_step, False, + sigma_ratio=sigma_factor, noise=noise, scheduler_func=scheduler_func) + else: + advanced_steps = math.floor(steps / denoise) + start_at_step = advanced_steps - steps + end_at_step = start_at_step + math.floor(steps * (1.0 - refiner_ratio)) + + # print(f"pre: {start_at_step} .. {end_at_step} / {advanced_steps}") + temp_latent = separated_sample(model, True, seed, advanced_steps, cfg, sampler_name, scheduler, + positive, negative, latent_image, start_at_step, end_at_step, True, + sigma_ratio=sigma_factor, noise=noise, scheduler_func=scheduler_func) + + if 'noise_mask' in latent_image: + # noise_latent = \ + # impact_sampling.separated_sample(refiner_model, "enable", seed, advanced_steps, cfg, sampler_name, + # scheduler, refiner_positive, refiner_negative, latent_image, end_at_step, + # end_at_step, "enable") + + latent_compositor = nodes.NODE_CLASS_MAPPINGS['LatentCompositeMasked']() + temp_latent = latent_compositor.composite(latent_image, temp_latent, 0, 0, False, latent_image['noise_mask'])[0] + + # print(f"post: {end_at_step} .. {advanced_steps + 1} / {advanced_steps}") + refined_latent = separated_sample(refiner_model, False, seed, advanced_steps, cfg, sampler_name, scheduler, + refiner_positive, refiner_negative, temp_latent, end_at_step, advanced_steps + 1, False, + sigma_ratio=sigma_factor, scheduler_func=scheduler_func) + + return refined_latent + + +class KSamplerAdvancedWrapper: + params = None + + def __init__(self, model, cfg, sampler_name, scheduler, positive, negative, sampler_opt=None, sigma_factor=1.0, scheduler_func=None): + self.params = model, cfg, sampler_name, scheduler, positive, negative, sigma_factor + self.sampler_opt = sampler_opt + self.scheduler_func = scheduler_func + + def clone_with_conditionings(self, positive, negative): + model, cfg, sampler_name, scheduler, _, _, _ = self.params + return KSamplerAdvancedWrapper(model, cfg, sampler_name, scheduler, positive, negative, self.sampler_opt) + + def sample_advanced(self, add_noise, seed, steps, latent_image, start_at_step, end_at_step, return_with_leftover_noise, hook=None, + recovery_mode="ratio additional", recovery_sampler="AUTO", recovery_sigma_ratio=1.0, noise=None): + + model, cfg, sampler_name, scheduler, positive, negative, sigma_factor = self.params + # steps, start_at_step, end_at_step = self.compensate_denoise(steps, start_at_step, end_at_step) + + if hook is not None: + model, seed, steps, cfg, sampler_name, scheduler, positive, negative, upscaled_latent = hook.pre_ksample_advanced(model, add_noise, seed, steps, cfg, sampler_name, scheduler, + positive, negative, latent_image, start_at_step, end_at_step, + return_with_leftover_noise) + + if recovery_mode != 'DISABLE' and sampler_name in ['uni_pc', 'uni_pc_bh2', 'dpmpp_sde', 'dpmpp_sde_gpu', 'dpmpp_2m_sde', 'dpmpp_2m_sde_gpu', 'dpmpp_3m_sde', 'dpmpp_3m_sde_gpu']: + base_image = latent_image.copy() + if recovery_mode == "ratio between": + sigma_ratio = 1.0 - recovery_sigma_ratio + else: + sigma_ratio = 1.0 + else: + base_image = None + sigma_ratio = 1.0 + + try: + if sigma_ratio > 0: + latent_image = separated_sample(model, add_noise, seed, steps, cfg, sampler_name, scheduler, + positive, negative, latent_image, start_at_step, end_at_step, + return_with_leftover_noise, sigma_ratio=sigma_ratio * sigma_factor, + sampler_opt=self.sampler_opt, noise=noise, scheduler_func=self.scheduler_func) + except ValueError as e: + if str(e) == 'sigma_min and sigma_max must not be 0': + print(f"\nWARN: sampling skipped - sigma_min and sigma_max are 0") + return latent_image + + if (recovery_sigma_ratio > 0 and recovery_mode != 'DISABLE' and + sampler_name in ['uni_pc', 'uni_pc_bh2', 'dpmpp_sde', 'dpmpp_sde_gpu', 'dpmpp_2m_sde', 'dpmpp_2m_sde_gpu', 'dpmpp_3m_sde', 'dpmpp_3m_sde_gpu']): + compensate = 0 if sampler_name in ['uni_pc', 'uni_pc_bh2', 'dpmpp_sde', 'dpmpp_sde_gpu', 'dpmpp_2m_sde', 'dpmpp_2m_sde_gpu', 'dpmpp_3m_sde', 'dpmpp_3m_sde_gpu'] else 2 + if recovery_sampler == "AUTO": + recovery_sampler = 'dpm_fast' if sampler_name in ['uni_pc', 'uni_pc_bh2', 'dpmpp_sde', 'dpmpp_sde_gpu'] else 'dpmpp_2m' + + latent_compositor = nodes.NODE_CLASS_MAPPINGS['LatentCompositeMasked']() + + noise_mask = latent_image['noise_mask'] + + if len(noise_mask.shape) == 4: + noise_mask = noise_mask.squeeze(0).squeeze(0) + + latent_image = latent_compositor.composite(base_image, latent_image, 0, 0, False, noise_mask)[0] + + try: + latent_image = separated_sample(model, add_noise, seed, steps, cfg, recovery_sampler, scheduler, + positive, negative, latent_image, start_at_step-compensate, end_at_step, return_with_leftover_noise, + sigma_ratio=recovery_sigma_ratio * sigma_factor, sampler_opt=self.sampler_opt, scheduler_func=self.scheduler_func) + except ValueError as e: + if str(e) == 'sigma_min and sigma_max must not be 0': + print(f"\nWARN: sampling skipped - sigma_min and sigma_max are 0") + + return latent_image + + +class KSamplerWrapper: + params = None + + def __init__(self, model, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise, scheduler_func=None): + self.params = model, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise + self.scheduler_func = scheduler_func + + def sample(self, latent_image, hook=None): + model, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise = self.params + + if hook is not None: + model, seed, steps, cfg, sampler_name, scheduler, positive, negative, upscaled_latent, denoise = \ + hook.pre_ksample(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise) + + return impact_sample(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise, scheduler_func=self.scheduler_func) diff --git a/ComfyUI-Impact-Pack/modules/impact/impact_server.py b/ComfyUI-Impact-Pack/modules/impact/impact_server.py new file mode 100644 index 0000000000000000000000000000000000000000..bd91d7a429199cef888d449d3b191b2ccf2a3785 --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/impact_server.py @@ -0,0 +1,541 @@ +import os +import threading +import traceback + +from aiohttp import web + +import impact +import folder_paths + +import torchvision + +import impact.core as core +import impact.impact_pack as impact_pack +from impact.utils import to_tensor +from segment_anything import SamPredictor, sam_model_registry +import numpy as np +import nodes +from PIL import Image +import io +import impact.wildcards as wildcards +import comfy +from io import BytesIO +import random +from server import PromptServer + + +sam_predictor = None +default_sam_model_name = os.path.join(impact_pack.model_path, "sams", "sam_vit_b_01ec64.pth") + +sam_lock = threading.Condition() + +last_prepare_data = None + + +def async_prepare_sam(image_dir, model_name, filename): + with sam_lock: + global sam_predictor + + if 'vit_h' in model_name: + model_kind = 'vit_h' + elif 'vit_l' in model_name: + model_kind = 'vit_l' + else: + model_kind = 'vit_b' + + sam_model = sam_model_registry[model_kind](checkpoint=model_name) + sam_predictor = SamPredictor(sam_model) + + image_path = os.path.join(image_dir, filename) + image = nodes.LoadImage().load_image(image_path)[0] + image = np.clip(255. * image.cpu().numpy().squeeze(), 0, 255).astype(np.uint8) + + if impact.config.get_config()['sam_editor_cpu']: + device = 'cpu' + else: + device = comfy.model_management.get_torch_device() + + sam_predictor.model.to(device=device) + sam_predictor.set_image(image, "RGB") + sam_predictor.model.cpu() + + +@PromptServer.instance.routes.post("/sam/prepare") +async def sam_prepare(request): + global sam_predictor + global last_prepare_data + data = await request.json() + + with sam_lock: + if last_prepare_data is not None and last_prepare_data == data: + # already loaded: skip -- prevent redundant loading + return web.Response(status=200) + + last_prepare_data = data + + model_name = 'sam_vit_b_01ec64.pth' + if data['sam_model_name'] == 'auto': + model_name = impact.config.get_config()['sam_editor_model'] + + model_name = os.path.join(impact_pack.model_path, "sams", model_name) + + print(f"[INFO] ComfyUI-Impact-Pack: Loading SAM model '{impact_pack.model_path}'") + + filename, image_dir = folder_paths.annotated_filepath(data["filename"]) + + if image_dir is None: + typ = data['type'] if data['type'] != '' else 'output' + image_dir = folder_paths.get_directory_by_type(typ) + if data['subfolder'] is not None and data['subfolder'] != '': + image_dir += f"/{data['subfolder']}" + + if image_dir is None: + return web.Response(status=400) + + thread = threading.Thread(target=async_prepare_sam, args=(image_dir, model_name, filename,)) + thread.start() + + print(f"[INFO] ComfyUI-Impact-Pack: SAM model loaded. ") + return web.Response(status=200) + + +@PromptServer.instance.routes.post("/sam/release") +async def release_sam(request): + global sam_predictor + + with sam_lock: + del sam_predictor + sam_predictor = None + + print(f"[INFO] ComfyUI-Impact-Pack: unloading SAM model") + + +@PromptServer.instance.routes.post("/sam/detect") +async def sam_detect(request): + global sam_predictor + with sam_lock: + if sam_predictor is not None: + if impact.config.get_config()['sam_editor_cpu']: + device = 'cpu' + else: + device = comfy.model_management.get_torch_device() + + sam_predictor.model.to(device=device) + try: + data = await request.json() + + positive_points = data['positive_points'] + negative_points = data['negative_points'] + threshold = data['threshold'] + + points = [] + plabs = [] + + for p in positive_points: + points.append(p) + plabs.append(1) + + for p in negative_points: + points.append(p) + plabs.append(0) + + detected_masks = core.sam_predict(sam_predictor, points, plabs, None, threshold) + mask = core.combine_masks2(detected_masks) + + if mask is None: + return web.Response(status=400) + + image = mask.reshape((-1, 1, mask.shape[-2], mask.shape[-1])).movedim(1, -1).expand(-1, -1, -1, 3) + i = 255. * image.cpu().numpy() + + img = Image.fromarray(np.clip(i[0], 0, 255).astype(np.uint8)) + + img_buffer = io.BytesIO() + img.save(img_buffer, format='png') + + headers = {'Content-Type': 'image/png'} + finally: + sam_predictor.model.to(device="cpu") + + return web.Response(body=img_buffer.getvalue(), headers=headers) + + else: + return web.Response(status=400) + + +@PromptServer.instance.routes.get("/impact/wildcards/refresh") +async def wildcards_refresh(request): + impact.wildcards.wildcard_load() + return web.Response(status=200) + + +@PromptServer.instance.routes.get("/impact/wildcards/list") +async def wildcards_list(request): + data = {'data': impact.wildcards.get_wildcard_list()} + return web.json_response(data) + + +@PromptServer.instance.routes.post("/impact/wildcards") +async def populate_wildcards(request): + data = await request.json() + populated = wildcards.process(data['text'], data.get('seed', None)) + return web.json_response({"text": populated}) + + +segs_picker_map = {} + +@PromptServer.instance.routes.get("/impact/segs/picker/count") +async def segs_picker_count(request): + node_id = request.rel_url.query.get('id', '') + + if node_id in segs_picker_map: + res = len(segs_picker_map[node_id]) + return web.Response(status=200, text=str(res)) + + return web.Response(status=400) + + +@PromptServer.instance.routes.get("/impact/segs/picker/view") +async def segs_picker(request): + node_id = request.rel_url.query.get('id', '') + idx = int(request.rel_url.query.get('idx', '')) + + if node_id in segs_picker_map and idx < len(segs_picker_map[node_id]): + img = to_tensor(segs_picker_map[node_id][idx]).permute(0, 3, 1, 2).squeeze(0) + pil = torchvision.transforms.ToPILImage('RGB')(img) + + image_bytes = BytesIO() + pil.save(image_bytes, format="PNG") + image_bytes.seek(0) + return web.Response(status=200, body=image_bytes, content_type='image/png', headers={"Content-Disposition": f"filename={node_id}{idx}.png"}) + + return web.Response(status=400) + + +@PromptServer.instance.routes.get("/view/validate") +async def view_validate(request): + if "filename" in request.rel_url.query: + filename = request.rel_url.query["filename"] + subfolder = request.rel_url.query["subfolder"] + filename, base_dir = folder_paths.annotated_filepath(filename) + + if filename == '' or filename[0] == '/' or '..' in filename: + return web.Response(status=400) + + if base_dir is None: + base_dir = folder_paths.get_input_directory() + + file = os.path.join(base_dir, subfolder, filename) + + if os.path.isfile(file): + return web.Response(status=200) + + return web.Response(status=400) + + +@PromptServer.instance.routes.get("/impact/validate/pb_id_image") +async def view_validate(request): + if "id" in request.rel_url.query: + pb_id = request.rel_url.query["id"] + + if pb_id not in core.preview_bridge_image_id_map: + return web.Response(status=400) + + file = core.preview_bridge_image_id_map[pb_id] + if os.path.isfile(file): + return web.Response(status=200) + + return web.Response(status=400) + + +@PromptServer.instance.routes.get("/impact/set/pb_id_image") +async def set_previewbridge_image(request): + try: + if "filename" in request.rel_url.query: + node_id = request.rel_url.query["node_id"] + filename = request.rel_url.query["filename"] + path_type = request.rel_url.query["type"] + subfolder = request.rel_url.query["subfolder"] + filename, output_dir = folder_paths.annotated_filepath(filename) + + if filename == '' or filename[0] == '/' or '..' in filename: + return web.Response(status=400) + + if output_dir is None: + if path_type == 'input': + output_dir = folder_paths.get_input_directory() + elif path_type == 'output': + output_dir = folder_paths.get_output_directory() + else: + output_dir = folder_paths.get_temp_directory() + + file = os.path.join(output_dir, subfolder, filename) + item = { + 'filename': filename, + 'type': path_type, + 'subfolder': subfolder, + } + pb_id = core.set_previewbridge_image(node_id, file, item) + + return web.Response(status=200, text=pb_id) + except Exception: + traceback.print_exc() + + return web.Response(status=400) + + +@PromptServer.instance.routes.get("/impact/get/pb_id_image") +async def get_previewbridge_image(request): + if "id" in request.rel_url.query: + pb_id = request.rel_url.query["id"] + + if pb_id in core.preview_bridge_image_id_map: + _, path_item = core.preview_bridge_image_id_map[pb_id] + return web.json_response(path_item) + + return web.Response(status=400) + + +@PromptServer.instance.routes.get("/impact/view/pb_id_image") +async def view_previewbridge_image(request): + if "id" in request.rel_url.query: + pb_id = request.rel_url.query["id"] + + if pb_id in core.preview_bridge_image_id_map: + file = core.preview_bridge_image_id_map[pb_id] + + with Image.open(file) as img: + filename = os.path.basename(file) + return web.FileResponse(file, headers={"Content-Disposition": f"filename=\"{filename}\""}) + + return web.Response(status=400) + + +def onprompt_for_switch(json_data): + inversed_switch_info = {} + onprompt_switch_info = {} + onprompt_cond_branch_info = {} + + for k, v in json_data['prompt'].items(): + if 'class_type' not in v: + continue + + cls = v['class_type'] + if cls == 'ImpactInversedSwitch': + if 'sel_mode' in v['inputs'] and v['inputs']['sel_mode'] and 'select' in v['inputs']: + select_input = v['inputs']['select'] + if isinstance(select_input, list) and len(select_input) == 2: + input_node = json_data['prompt'][select_input[0]] + if input_node['class_type'] == 'ImpactInt' and 'inputs' in input_node and 'value' in input_node['inputs']: + inversed_switch_info[k] = input_node['inputs']['value'] + else: + print(f"\n##### ##### #####\n[WARN] {cls}: For the 'select' operation, only 'select_index' of the 'ImpactInversedSwitch', which is not an input, or 'ImpactInt' and 'Primitive' are allowed as inputs if 'select_on_prompt' is selected.\n##### ##### #####\n") + else: + inversed_switch_info[k] = select_input + + elif cls in ['ImpactSwitch', 'LatentSwitch', 'SEGSSwitch', 'ImpactMakeImageList']: + if 'sel_mode' in v['inputs'] and v['inputs']['sel_mode'] and 'select' in v['inputs']: + select_input = v['inputs']['select'] + if isinstance(select_input, list) and len(select_input) == 2: + input_node = json_data['prompt'][select_input[0]] + if input_node['class_type'] == 'ImpactInt' and 'inputs' in input_node and 'value' in input_node['inputs']: + onprompt_switch_info[k] = input_node['inputs']['value'] + if input_node['class_type'] == 'ImpactSwitch' and 'inputs' in input_node and 'select' in input_node['inputs']: + if isinstance(input_node['inputs']['select'], int): + onprompt_switch_info[k] = input_node['inputs']['select'] + else: + print(f"\n##### ##### #####\n[WARN] {cls}: For the 'select' operation, only 'select_index' of the 'ImpactSwitch', which is not an input, or 'ImpactInt' and 'Primitive' are allowed as inputs if 'select_on_prompt' is selected.\n##### ##### #####\n") + else: + onprompt_switch_info[k] = select_input + + elif cls == 'ImpactConditionalBranchSelMode': + if 'sel_mode' in v['inputs'] and v['inputs']['sel_mode'] and 'cond' in v['inputs']: + cond_input = v['inputs']['cond'] + if isinstance(cond_input, list) and len(cond_input) == 2: + input_node = json_data['prompt'][cond_input[0]] + if (input_node['class_type'] == 'ImpactValueReceiver' and 'inputs' in input_node + and 'value' in input_node['inputs'] and 'typ' in input_node['inputs']): + if 'BOOLEAN' == input_node['inputs']['typ']: + try: + onprompt_cond_branch_info[k] = input_node['inputs']['value'].lower() == "true" + except: + pass + else: + onprompt_cond_branch_info[k] = cond_input + + for k, v in json_data['prompt'].items(): + disable_targets = set() + + for kk, vv in v['inputs'].items(): + if isinstance(vv, list) and len(vv) == 2: + if vv[0] in inversed_switch_info: + if vv[1] + 1 != inversed_switch_info[vv[0]]: + disable_targets.add(kk) + + if k in onprompt_switch_info: + selected_slot_name = f"input{onprompt_switch_info[k]}" + for kk, vv in v['inputs'].items(): + if kk != selected_slot_name and kk.startswith('input'): + disable_targets.add(kk) + + if k in onprompt_cond_branch_info: + selected_slot_name = "tt_value" if onprompt_cond_branch_info[k] else "ff_value" + for kk, vv in v['inputs'].items(): + if kk in ['tt_value', 'ff_value'] and kk != selected_slot_name: + disable_targets.add(kk) + + for kk in disable_targets: + del v['inputs'][kk] + +def onprompt_for_pickers(json_data): + detected_pickers = set() + + for k, v in json_data['prompt'].items(): + if 'class_type' not in v: + continue + + cls = v['class_type'] + if cls == 'ImpactSEGSPicker': + detected_pickers.add(k) + + # garbage collection + keys_to_remove = [key for key in segs_picker_map if key not in detected_pickers] + for key in keys_to_remove: + del segs_picker_map[key] + + +def gc_preview_bridge_cache(json_data): + prompt_keys = json_data['prompt'].keys() + + for key in list(core.preview_bridge_cache.keys()): + if key not in prompt_keys: + # print(f"key deleted [PB]: {key}") + del core.preview_bridge_cache[key] + + for key in list(core.preview_bridge_last_mask_cache.keys()): + if key not in prompt_keys: + # print(f"key deleted [PB_last_mask]: {key}") + del core.preview_bridge_last_mask_cache[key] + + +def workflow_imagereceiver_update(json_data): + prompt = json_data['prompt'] + + for v in prompt.values(): + if 'class_type' in v and v['class_type'] == 'ImageReceiver': + if v['inputs']['save_to_workflow']: + v['inputs']['image'] = "#DATA" + + +def regional_sampler_seed_update(json_data): + prompt = json_data['prompt'] + + for k, v in prompt.items(): + if 'class_type' in v and v['class_type'] == 'RegionalSampler': + seed_2nd_mode = v['inputs']['seed_2nd_mode'] + + new_seed = None + if seed_2nd_mode == 'increment': + new_seed = v['inputs']['seed_2nd']+1 + if new_seed > 1125899906842624: + new_seed = 0 + elif seed_2nd_mode == 'decrement': + new_seed = v['inputs']['seed_2nd']-1 + if new_seed < 0: + new_seed = 1125899906842624 + elif seed_2nd_mode == 'randomize': + new_seed = random.randint(0, 1125899906842624) + + if new_seed is not None: + PromptServer.instance.send_sync("impact-node-feedback", {"node_id": k, "widget_name": "seed_2nd", "type": "INT", "value": new_seed}) + + +def onprompt_populate_wildcards(json_data): + prompt = json_data['prompt'] + + updated_widget_values = {} + for k, v in prompt.items(): + if 'class_type' in v and (v['class_type'] == 'ImpactWildcardEncode' or v['class_type'] == 'ImpactWildcardProcessor'): + inputs = v['inputs'] + if inputs['mode'] and isinstance(inputs['populated_text'], str): + if isinstance(inputs['seed'], list): + try: + input_node = prompt[inputs['seed'][0]] + if input_node['class_type'] == 'ImpactInt': + input_seed = int(input_node['inputs']['value']) + if not isinstance(input_seed, int): + continue + if input_node['class_type'] == 'Seed (rgthree)': + input_seed = int(input_node['inputs']['seed']) + if not isinstance(input_seed, int): + continue + else: + print(f"[Impact Pack] Only `ImpactInt`, `Seed (rgthree)` and `Primitive` Node are allowed as the seed for '{v['class_type']}'. It will be ignored. ") + continue + except: + continue + else: + input_seed = int(inputs['seed']) + + inputs['populated_text'] = wildcards.process(inputs['wildcard_text'], input_seed) + inputs['mode'] = False + + PromptServer.instance.send_sync("impact-node-feedback", {"node_id": k, "widget_name": "populated_text", "type": "STRING", "value": inputs['populated_text']}) + updated_widget_values[k] = inputs['populated_text'] + + if 'extra_data' in json_data and 'extra_pnginfo' in json_data['extra_data']: + for node in json_data['extra_data']['extra_pnginfo']['workflow']['nodes']: + key = str(node['id']) + if key in updated_widget_values: + node['widgets_values'][1] = updated_widget_values[key] + node['widgets_values'][2] = False + + +def onprompt_for_remote(json_data): + prompt = json_data['prompt'] + + for v in prompt.values(): + if 'class_type' in v: + cls = v['class_type'] + if cls == 'ImpactRemoteBoolean' or cls == 'ImpactRemoteInt': + inputs = v['inputs'] + node_id = str(inputs['node_id']) + + if node_id not in prompt: + continue + + target_inputs = prompt[node_id]['inputs'] + + widget_name = inputs['widget_name'] + if widget_name in target_inputs: + widget_type = None + if cls == 'ImpactRemoteBoolean' and isinstance(target_inputs[widget_name], bool): + widget_type = 'BOOLEAN' + + elif cls == 'ImpactRemoteInt' and (isinstance(target_inputs[widget_name], int) or isinstance(target_inputs[widget_name], float)): + widget_type = 'INT' + + if widget_type is None: + break + + target_inputs[widget_name] = inputs['value'] + PromptServer.instance.send_sync("impact-node-feedback", {"node_id": node_id, "widget_name": widget_name, "type": widget_type, "value": inputs['value']}) + + +def onprompt(json_data): + try: + onprompt_for_remote(json_data) # NOTE: top priority + onprompt_for_switch(json_data) + onprompt_for_pickers(json_data) + onprompt_populate_wildcards(json_data) + gc_preview_bridge_cache(json_data) + workflow_imagereceiver_update(json_data) + regional_sampler_seed_update(json_data) + core.current_prompt = json_data + except Exception as e: + print(f"[WARN] ComfyUI-Impact-Pack: Error on prompt - several features will not work.\n{e}") + + return json_data + + +PromptServer.instance.add_on_prompt_handler(onprompt) diff --git a/ComfyUI-Impact-Pack/modules/impact/legacy_nodes.py b/ComfyUI-Impact-Pack/modules/impact/legacy_nodes.py new file mode 100644 index 0000000000000000000000000000000000000000..d7e10aff8c9c50628b23fcd717d0099987620519 --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/legacy_nodes.py @@ -0,0 +1,285 @@ +import folder_paths + +import impact.mmdet_nodes as mmdet_nodes +from impact.utils import * +from impact.core import SEG +import impact.core as core +import nodes + +class NO_BBOX_MODEL: + pass + + +class NO_SEGM_MODEL: + pass + + +class MMDetLoader: + @classmethod + def INPUT_TYPES(s): + bboxs = ["bbox/"+x for x in folder_paths.get_filename_list("mmdets_bbox")] + segms = ["segm/"+x for x in folder_paths.get_filename_list("mmdets_segm")] + return {"required": {"model_name": (bboxs + segms, )}} + RETURN_TYPES = ("BBOX_MODEL", "SEGM_MODEL") + FUNCTION = "load_mmdet" + + CATEGORY = "ImpactPack/Legacy" + + DEPRECATED = True + + def load_mmdet(self, model_name): + mmdet_path = folder_paths.get_full_path("mmdets", model_name) + model = mmdet_nodes.load_mmdet(mmdet_path) + + if model_name.startswith("bbox"): + return model, NO_SEGM_MODEL() + else: + return NO_BBOX_MODEL(), model + + +class BboxDetectorForEach: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "bbox_model": ("BBOX_MODEL", ), + "image": ("IMAGE", ), + "threshold": ("FLOAT", {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01}), + "dilation": ("INT", {"default": 10, "min": 0, "max": 255, "step": 1}), + "crop_factor": ("FLOAT", {"default": 3.0, "min": 1.0, "max": 100, "step": 0.1}), + } + } + + RETURN_TYPES = ("SEGS", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Legacy" + + DEPRECATED = True + + @staticmethod + def detect(bbox_model, image, threshold, dilation, crop_factor, drop_size=1, detailer_hook=None): + mmdet_results = mmdet_nodes.inference_bbox(bbox_model, image, threshold) + segmasks = core.create_segmasks(mmdet_results) + + if dilation > 0: + segmasks = dilate_masks(segmasks, dilation) + + items = [] + h = image.shape[1] + w = image.shape[2] + for x in segmasks: + item_bbox = x[0] + item_mask = x[1] + + y1, x1, y2, x2 = item_bbox + + if x2 - x1 > drop_size and y2 - y1 > drop_size: + crop_region = make_crop_region(w, h, item_bbox, crop_factor) + cropped_image = crop_image(image, crop_region) + cropped_mask = crop_ndarray2(item_mask, crop_region) + confidence = x[2] + # bbox_size = (item_bbox[2]-item_bbox[0],item_bbox[3]-item_bbox[1]) # (w,h) + + item = SEG(cropped_image, cropped_mask, confidence, crop_region, item_bbox, None, None) + items.append(item) + + shape = h, w + return shape, items + + def doit(self, bbox_model, image, threshold, dilation, crop_factor): + return (BboxDetectorForEach.detect(bbox_model, image, threshold, dilation, crop_factor), ) + + +class SegmDetectorCombined: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segm_model": ("SEGM_MODEL", ), + "image": ("IMAGE", ), + "threshold": ("FLOAT", {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01}), + "dilation": ("INT", {"default": 0, "min": 0, "max": 255, "step": 1}), + } + } + + RETURN_TYPES = ("MASK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Legacy" + + DEPRECATED = True + + def doit(self, segm_model, image, threshold, dilation): + mmdet_results = mmdet_nodes.inference_segm(image, segm_model, threshold) + segmasks = core.create_segmasks(mmdet_results) + if dilation > 0: + segmasks = dilate_masks(segmasks, dilation) + + mask = combine_masks(segmasks) + return (mask,) + + +class BboxDetectorCombined(SegmDetectorCombined): + @classmethod + def INPUT_TYPES(s): + return {"required": { + "bbox_model": ("BBOX_MODEL", ), + "image": ("IMAGE", ), + "threshold": ("FLOAT", {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01}), + "dilation": ("INT", {"default": 4, "min": 0, "max": 255, "step": 1}), + } + } + + def doit(self, bbox_model, image, threshold, dilation): + mmdet_results = mmdet_nodes.inference_bbox(bbox_model, image, threshold) + segmasks = core.create_segmasks(mmdet_results) + if dilation > 0: + segmasks = dilate_masks(segmasks, dilation) + + mask = combine_masks(segmasks) + return (mask,) + + +class SegmDetectorForEach: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segm_model": ("SEGM_MODEL", ), + "image": ("IMAGE", ), + "threshold": ("FLOAT", {"default": 0.5, "min": 0.0, "max": 1.0, "step": 0.01}), + "dilation": ("INT", {"default": 10, "min": 0, "max": 255, "step": 1}), + "crop_factor": ("FLOAT", {"default": 3.0, "min": 1.0, "max": 100, "step": 0.1}), + } + } + + RETURN_TYPES = ("SEGS", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Legacy" + + DEPRECATED = True + + def doit(self, segm_model, image, threshold, dilation, crop_factor): + mmdet_results = mmdet_nodes.inference_segm(image, segm_model, threshold) + segmasks = core.create_segmasks(mmdet_results) + + if dilation > 0: + segmasks = dilate_masks(segmasks, dilation) + + items = [] + h = image.shape[1] + w = image.shape[2] + for x in segmasks: + item_bbox = x[0] + item_mask = x[1] + + crop_region = make_crop_region(w, h, item_bbox, crop_factor) + cropped_image = crop_image(image, crop_region) + cropped_mask = crop_ndarray2(item_mask, crop_region) + confidence = x[2] + + item = SEG(cropped_image, cropped_mask, confidence, crop_region, item_bbox, None, None) + items.append(item) + + shape = h,w + return ((shape, items), ) + + +class SegsMaskCombine: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS", ), + "image": ("IMAGE", ), + } + } + + RETURN_TYPES = ("MASK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Legacy" + + DEPRECATED = True + + @staticmethod + def combine(segs, image): + h = image.shape[1] + w = image.shape[2] + + mask = np.zeros((h, w), dtype=np.uint8) + + for seg in segs[1]: + cropped_mask = seg.cropped_mask + crop_region = seg.crop_region + mask[crop_region[1]:crop_region[3], crop_region[0]:crop_region[2]] |= (cropped_mask * 255).astype(np.uint8) + + return torch.from_numpy(mask.astype(np.float32) / 255.0) + + def doit(self, segs, image): + return (SegsMaskCombine.combine(segs, image), ) + + +class MaskPainter(nodes.PreviewImage): + @classmethod + def INPUT_TYPES(s): + return {"required": {"images": ("IMAGE",), }, + "hidden": { + "prompt": "PROMPT", + "extra_pnginfo": "EXTRA_PNGINFO", + }, + "optional": {"mask_image": ("IMAGE_PATH",), }, + "optional": {"image": (["#placeholder"], )}, + } + + RETURN_TYPES = ("MASK",) + + FUNCTION = "save_painted_images" + + CATEGORY = "ImpactPack/Legacy" + + DEPRECATED = True + + def save_painted_images(self, images, filename_prefix="impact-mask", + prompt=None, extra_pnginfo=None, mask_image=None, image=None): + if image == "#placeholder" or image['image_hash'] != id(images): + # new input image + res = self.save_images(images, filename_prefix, prompt, extra_pnginfo) + + item = res['ui']['images'][0] + + if not item['filename'].endswith(']'): + filepath = f"{item['filename']} [{item['type']}]" + else: + filepath = item['filename'] + + _, mask = nodes.LoadImage().load_image(filepath) + + res['ui']['aux'] = [id(images), res['ui']['images']] + res['result'] = (mask, ) + + return res + + else: + # new mask + if '0' in image: # fallback + image = image['0'] + + forward = {'filename': image['forward_filename'], + 'subfolder': image['forward_subfolder'], + 'type': image['forward_type'], } + + res = {'ui': {'images': [forward]}} + + imgpath = "" + if 'subfolder' in image and image['subfolder'] != "": + imgpath = image['subfolder'] + "/" + + imgpath += f"{image['filename']}" + + if 'type' in image and image['type'] != "": + imgpath += f" [{image['type']}]" + + res['ui']['aux'] = [id(images), [forward]] + _, mask = nodes.LoadImage().load_image(imgpath) + res['result'] = (mask, ) + + return res diff --git a/ComfyUI-Impact-Pack/modules/impact/logics.py b/ComfyUI-Impact-Pack/modules/impact/logics.py new file mode 100644 index 0000000000000000000000000000000000000000..e57ff8b008de299f0a129244cced318dcdb99d7c --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/logics.py @@ -0,0 +1,774 @@ +import sys +import time + +import execution +import impact.impact_server +from server import PromptServer +from impact.utils import any_typ +import impact.core as core +import re +import nodes +import traceback + +class ImpactCompare: + @classmethod + def INPUT_TYPES(cls): + return { + "required": { + "cmp": (['a = b', 'a <> b', 'a > b', 'a < b', 'a >= b', 'a <= b', 'tt', 'ff'],), + "a": (any_typ, ), + "b": (any_typ, ), + }, + } + + FUNCTION = "doit" + CATEGORY = "ImpactPack/Logic" + + RETURN_TYPES = ("BOOLEAN", ) + + def doit(self, cmp, a, b): + if cmp == "a = b": + return (a == b, ) + elif cmp == "a <> b": + return (a != b, ) + elif cmp == "a > b": + return (a > b, ) + elif cmp == "a < b": + return (a < b, ) + elif cmp == "a >= b": + return (a >= b, ) + elif cmp == "a <= b": + return (a <= b, ) + elif cmp == 'tt': + return (True, ) + else: + return (False, ) + + +class ImpactNotEmptySEGS: + @classmethod + def INPUT_TYPES(cls): + return {"required": {"segs": ("SEGS",)}} + + FUNCTION = "doit" + CATEGORY = "ImpactPack/Logic" + + RETURN_TYPES = ("BOOLEAN", ) + + def doit(self, segs): + return (segs[1] != [], ) + + +class ImpactConditionalBranch: + @classmethod + def INPUT_TYPES(cls): + return { + "required": { + "cond": ("BOOLEAN",), + "tt_value": (any_typ,{"lazy": True}), + "ff_value": (any_typ,{"lazy": True}), + }, + } + + FUNCTION = "doit" + CATEGORY = "ImpactPack/Logic" + + RETURN_TYPES = (any_typ, ) + + def check_lazy_status(self, cond, tt_value=None, ff_value=None): + if cond and tt_value is None: + return ["tt_value"] + if not cond and ff_value is None: + return ["ff_value"] + + def doit(self, cond, tt_value=None, ff_value=None): + if cond: + return (tt_value,) + else: + return (ff_value,) + + +class ImpactConditionalBranchSelMode: + @classmethod + def INPUT_TYPES(cls): + if not core.is_execution_model_version_supported(): + required_inputs = { + "cond": ("BOOLEAN",), + "sel_mode": ("BOOLEAN", {"default": True, "label_on": "select_on_prompt", "label_off": "select_on_execution"}), + } + else: + required_inputs = { + "cond": ("BOOLEAN",), + } + + return { + "required": required_inputs, + "optional": { + "tt_value": (any_typ,), + "ff_value": (any_typ,), + }, + } + + FUNCTION = "doit" + CATEGORY = "ImpactPack/Logic" + + RETURN_TYPES = (any_typ, ) + + def doit(self, cond, tt_value=None, ff_value=None, **kwargs): + print(f'tt={tt_value is None}\nff={ff_value is None}') + if cond: + return (tt_value,) + else: + return (ff_value,) + + +class ImpactConvertDataType: + def __init__(self): + pass + + @classmethod + def INPUT_TYPES(cls): + return {"required": {"value": (any_typ,)}} + + RETURN_TYPES = ("STRING", "FLOAT", "INT", "BOOLEAN") + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Logic" + + @staticmethod + def is_number(string): + pattern = re.compile(r'^[-+]?[0-9]*\.?[0-9]+$') + return bool(pattern.match(string)) + + def doit(self, value): + if self.is_number(str(value)): + num = value + else: + if str.lower(str(value)) != "false": + num = 1 + else: + num = 0 + return (str(value), float(num), int(float(num)), bool(float(num)), ) + + +class ImpactIfNone: + def __init__(self): + pass + + @classmethod + def INPUT_TYPES(cls): + return { + "required": {}, + "optional": {"signal": (any_typ,), "any_input": (any_typ,), } + } + + RETURN_TYPES = (any_typ, "BOOLEAN") + RETURN_NAMES = ("signal_opt", "bool") + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Logic" + + def doit(self, signal=None, any_input=None): + if any_input is None: + return (signal, False, ) + else: + return (signal, True, ) + + +class ImpactLogicalOperators: + @classmethod + def INPUT_TYPES(cls): + return { + "required": { + "operator": (['and', 'or', 'xor'],), + "bool_a": ("BOOLEAN", {"forceInput": True}), + "bool_b": ("BOOLEAN", {"forceInput": True}), + }, + } + + FUNCTION = "doit" + CATEGORY = "ImpactPack/Logic" + + RETURN_TYPES = ("BOOLEAN", ) + + def doit(self, operator, bool_a, bool_b): + if operator == "and": + return (bool_a and bool_b, ) + elif operator == "or": + return (bool_a or bool_b, ) + else: + return (bool_a != bool_b, ) + + +class ImpactConditionalStopIteration: + @classmethod + def INPUT_TYPES(cls): + return { + "required": { "cond": ("BOOLEAN", {"forceInput": True}), }, + } + + FUNCTION = "doit" + CATEGORY = "ImpactPack/Logic" + + RETURN_TYPES = () + + OUTPUT_NODE = True + + def doit(self, cond): + if cond: + PromptServer.instance.send_sync("stop-iteration", {}) + return {} + + +class ImpactNeg: + @classmethod + def INPUT_TYPES(cls): + return { + "required": { "value": ("BOOLEAN", {"forceInput": True}), }, + } + + FUNCTION = "doit" + CATEGORY = "ImpactPack/Logic" + + RETURN_TYPES = ("BOOLEAN", ) + + def doit(self, value): + return (not value, ) + + +class ImpactInt: + @classmethod + def INPUT_TYPES(cls): + return { + "required": { + "value": ("INT", {"default": 0, "min": 0, "max": sys.maxsize, "step": 1}), + }, + } + + FUNCTION = "doit" + CATEGORY = "ImpactPack/Logic" + + RETURN_TYPES = ("INT", ) + + def doit(self, value): + return (value, ) + + +class ImpactFloat: + @classmethod + def INPUT_TYPES(cls): + return { + "required": { + "value": ("FLOAT", {"default": 1.0, "min": -3.402823466e+38, "max": 3.402823466e+38}), + }, + } + + FUNCTION = "doit" + CATEGORY = "ImpactPack/Logic" + + RETURN_TYPES = ("FLOAT", ) + + def doit(self, value): + return (value, ) + + +class ImpactBoolean: + @classmethod + def INPUT_TYPES(cls): + return { + "required": { + "value": ("BOOLEAN", {"default": False}), + }, + } + + FUNCTION = "doit" + CATEGORY = "ImpactPack/Logic" + + RETURN_TYPES = ("BOOLEAN", ) + + def doit(self, value): + return (value, ) + + +class ImpactValueSender: + @classmethod + def INPUT_TYPES(cls): + return {"required": { + "value": (any_typ, ), + "link_id": ("INT", {"default": 0, "min": 0, "max": sys.maxsize, "step": 1}), + }, + "optional": { + "signal_opt": (any_typ,), + } + } + + OUTPUT_NODE = True + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Logic" + + RETURN_TYPES = (any_typ, ) + RETURN_NAMES = ("signal", ) + + def doit(self, value, link_id=0, signal_opt=None): + PromptServer.instance.send_sync("value-send", {"link_id": link_id, "value": value}) + return (signal_opt, ) + + +class ImpactIntConstSender: + @classmethod + def INPUT_TYPES(cls): + return {"required": { + "signal": (any_typ, ), + "value": ("INT", {"default": 0, "min": 0, "max": sys.maxsize, "step": 1}), + "link_id": ("INT", {"default": 0, "min": 0, "max": sys.maxsize, "step": 1}), + }, + } + + OUTPUT_NODE = True + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Logic" + + RETURN_TYPES = () + + def doit(self, signal, value, link_id=0): + PromptServer.instance.send_sync("value-send", {"link_id": link_id, "value": value}) + return {} + + +class ImpactValueReceiver: + @classmethod + def INPUT_TYPES(cls): + return {"required": { + "typ": (["STRING", "INT", "FLOAT", "BOOLEAN"], ), + "value": ("STRING", {"default": ""}), + "link_id": ("INT", {"default": 0, "min": 0, "max": sys.maxsize, "step": 1}), + }, + } + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Logic" + + RETURN_TYPES = (any_typ, ) + + def doit(self, typ, value, link_id=0): + if typ == "INT": + return (int(value), ) + elif typ == "FLOAT": + return (float(value), ) + elif typ == "BOOLEAN": + return (value.lower() == "true", ) + else: + return (value, ) + + +class ImpactImageInfo: + @classmethod + def INPUT_TYPES(cls): + return {"required": { + "value": ("IMAGE", ), + }, + } + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Logic/_for_test" + + RETURN_TYPES = ("INT", "INT", "INT", "INT") + RETURN_NAMES = ("batch", "height", "width", "channel") + + def doit(self, value): + return (value.shape[0], value.shape[1], value.shape[2], value.shape[3]) + + +class ImpactLatentInfo: + @classmethod + def INPUT_TYPES(cls): + return {"required": { + "value": ("LATENT", ), + }, + } + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Logic/_for_test" + + RETURN_TYPES = ("INT", "INT", "INT", "INT") + RETURN_NAMES = ("batch", "height", "width", "channel") + + def doit(self, value): + shape = value['samples'].shape + return (shape[0], shape[2] * 8, shape[3] * 8, shape[1]) + + +class ImpactMinMax: + @classmethod + def INPUT_TYPES(cls): + return {"required": { + "mode": ("BOOLEAN", {"default": True, "label_on": "max", "label_off": "min"}), + "a": (any_typ,), + "b": (any_typ,), + }, + } + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Logic/_for_test" + + RETURN_TYPES = ("INT", ) + + def doit(self, mode, a, b): + if mode: + return (max(a, b), ) + else: + return (min(a, b),) + + +class ImpactQueueTrigger: + @classmethod + def INPUT_TYPES(cls): + return {"required": { + "signal": (any_typ,), + "mode": ("BOOLEAN", {"default": True, "label_on": "Trigger", "label_off": "Don't trigger"}), + } + } + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Logic/_for_test" + RETURN_TYPES = (any_typ,) + RETURN_NAMES = ("signal_opt",) + OUTPUT_NODE = True + + def doit(self, signal, mode): + if(mode): + PromptServer.instance.send_sync("impact-add-queue", {}) + + return (signal,) + + +class ImpactQueueTriggerCountdown: + @classmethod + def INPUT_TYPES(cls): + return {"required": { + "count": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "total": ("INT", {"default": 10, "min": 1, "max": 0xffffffffffffffff}), + "mode": ("BOOLEAN", {"default": True, "label_on": "Trigger", "label_off": "Don't trigger"}), + }, + "optional": {"signal": (any_typ,),}, + "hidden": {"unique_id": "UNIQUE_ID"} + } + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Logic/_for_test" + RETURN_TYPES = (any_typ, "INT", "INT") + RETURN_NAMES = ("signal_opt", "count", "total") + OUTPUT_NODE = True + + def doit(self, count, total, mode, unique_id, signal=None): + if (mode): + if count < total - 1: + PromptServer.instance.send_sync("impact-node-feedback", + {"node_id": unique_id, "widget_name": "count", "type": "int", "value": count+1}) + PromptServer.instance.send_sync("impact-add-queue", {}) + if count >= total - 1: + PromptServer.instance.send_sync("impact-node-feedback", + {"node_id": unique_id, "widget_name": "count", "type": "int", "value": 0}) + + return (signal, count, total) + + + +class ImpactSetWidgetValue: + @classmethod + def INPUT_TYPES(cls): + return {"required": { + "signal": (any_typ,), + "node_id": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "widget_name": ("STRING", {"multiline": False}), + }, + "optional": { + "boolean_value": ("BOOLEAN", {"forceInput": True}), + "int_value": ("INT", {"forceInput": True}), + "float_value": ("FLOAT", {"forceInput": True}), + "string_value": ("STRING", {"forceInput": True}), + } + } + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Logic/_for_test" + RETURN_TYPES = (any_typ,) + RETURN_NAMES = ("signal_opt",) + OUTPUT_NODE = True + + def doit(self, signal, node_id, widget_name, boolean_value=None, int_value=None, float_value=None, string_value=None, ): + kind = None + if boolean_value is not None: + value = boolean_value + kind = "BOOLEAN" + elif int_value is not None: + value = int_value + kind = "INT" + elif float_value is not None: + value = float_value + kind = "FLOAT" + elif string_value is not None: + value = string_value + kind = "STRING" + else: + value = None + + if value is not None: + PromptServer.instance.send_sync("impact-node-feedback", + {"node_id": node_id, "widget_name": widget_name, "type": kind, "value": value}) + + return (signal,) + + +class ImpactNodeSetMuteState: + @classmethod + def INPUT_TYPES(cls): + return {"required": { + "signal": (any_typ,), + "node_id": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "set_state": ("BOOLEAN", {"default": True, "label_on": "active", "label_off": "mute"}), + } + } + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Logic/_for_test" + RETURN_TYPES = (any_typ,) + RETURN_NAMES = ("signal_opt",) + OUTPUT_NODE = True + + def doit(self, signal, node_id, set_state): + PromptServer.instance.send_sync("impact-node-mute-state", {"node_id": node_id, "is_active": set_state}) + return (signal,) + + +class ImpactSleep: + @classmethod + def INPUT_TYPES(cls): + return {"required": { + "signal": (any_typ,), + "seconds": ("FLOAT", {"default": 0.5, "min": 0, "max": 3600}), + } + } + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Logic/_for_test" + RETURN_TYPES = (any_typ,) + RETURN_NAMES = ("signal_opt",) + OUTPUT_NODE = True + + def doit(self, signal, seconds): + time.sleep(seconds) + return (signal,) + + +error_skip_flag = False +try: + import cm_global + def filter_message(str): + global error_skip_flag + + if "IMPACT-PACK-SIGNAL: STOP CONTROL BRIDGE" in str: + return True + elif error_skip_flag and "ERROR:root:!!! Exception during processing !!!\n" == str: + error_skip_flag = False + return True + else: + return False + + cm_global.try_call(api='cm.register_message_collapse', f=filter_message) + +except Exception as e: + print(f"[WARN] ComfyUI-Impact-Pack: `ComfyUI` or `ComfyUI-Manager` is an outdated version.") + pass + + +def workflow_to_map(workflow): + nodes = {} + links = {} + for link in workflow['links']: + links[link[0]] = link[1:] + for node in workflow['nodes']: + nodes[str(node['id'])] = node + + return nodes, links + + +class ImpactRemoteBoolean: + @classmethod + def INPUT_TYPES(cls): + return {"required": { + "node_id": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "widget_name": ("STRING", {"multiline": False}), + "value": ("BOOLEAN", {"default": True, "label_on": "True", "label_off": "False"}), + }} + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Logic/_for_test" + RETURN_TYPES = () + OUTPUT_NODE = True + + def doit(self, **kwargs): + return {} + + +class ImpactRemoteInt: + @classmethod + def INPUT_TYPES(cls): + return {"required": { + "node_id": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "widget_name": ("STRING", {"multiline": False}), + "value": ("INT", {"default": 0, "min": -0xffffffffffffffff, "max": 0xffffffffffffffff}), + }} + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Logic/_for_test" + RETURN_TYPES = () + OUTPUT_NODE = True + + def doit(self, **kwargs): + return {} + +class ImpactControlBridge: + @classmethod + def INPUT_TYPES(cls): + return {"required": { + "value": (any_typ,), + "mode": ("BOOLEAN", {"default": True, "label_on": "Active", "label_off": "Stop/Mute/Bypass"}), + "behavior": (["Stop", "Mute", "Bypass"], ), + }, + "hidden": {"unique_id": "UNIQUE_ID", "prompt": "PROMPT", "extra_pnginfo": "EXTRA_PNGINFO"} + } + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Logic" + RETURN_TYPES = (any_typ,) + RETURN_NAMES = ("value",) + OUTPUT_NODE = True + + DESCRIPTION = ("When behavior is Stop and mode is active, the input value is passed directly to the output.\n" + "When behavior is Mute/Bypass and mode is active, the node connected to the output is changed to active state.\n" + "When behavior is Stop and mode is Stop/Mute/Bypass, the workflow execution of the current node is halted.\n" + "When behavior is Mute/Bypass and mode is Stop/Mute/Bypass, the node connected to the output is changed to Mute/Bypass state.") + + @classmethod + def IS_CHANGED(self, value, mode, behavior="Stop", unique_id=None, prompt=None, extra_pnginfo=None): + if behavior == "Stop": + return value, mode, behavior + else: + # NOTE: extra_pnginfo is not populated for IS_CHANGED. + # so extra_pnginfo is useless in here + try: + workflow = core.current_prompt['extra_data']['extra_pnginfo']['workflow'] + except: + print(f"[Impact Pack] core.current_prompt['extra_data']['extra_pnginfo']['workflow']") + return 0 + + nodes, links = workflow_to_map(workflow) + next_nodes = [] + + for link in nodes[unique_id]['outputs'][0]['links']: + node_id = str(links[link][2]) + impact.utils.collect_non_reroute_nodes(nodes, links, next_nodes, node_id) + + return next_nodes + + def doit(self, value, mode, behavior="Stop", unique_id=None, prompt=None, extra_pnginfo=None): + global error_skip_flag + + if core.is_execution_model_version_supported(): + from comfy_execution.graph import ExecutionBlocker + else: + print("[Impact Pack] ImpactControlBridge: ComfyUI is outdated. The 'Stop' behavior cannot function properly.") + + if behavior == "Stop": + if mode: + return (value, ) + else: + return (ExecutionBlocker(None), ) + else: + workflow_nodes, links = workflow_to_map(extra_pnginfo['workflow']) + + active_nodes = [] + mute_nodes = [] + bypass_nodes = [] + + for link in workflow_nodes[unique_id]['outputs'][0]['links']: + node_id = str(links[link][2]) + + next_nodes = [] + impact.utils.collect_non_reroute_nodes(workflow_nodes, links, next_nodes, node_id) + + for next_node_id in next_nodes: + node_mode = workflow_nodes[next_node_id]['mode'] + + if node_mode == 0: + active_nodes.append(next_node_id) + elif node_mode == 2: + mute_nodes.append(next_node_id) + elif node_mode == 4: + bypass_nodes.append(next_node_id) + + if mode: + # active + should_be_active_nodes = mute_nodes + bypass_nodes + if len(should_be_active_nodes) > 0: + PromptServer.instance.send_sync("impact-bridge-continue", {"node_id": unique_id, 'actives': list(should_be_active_nodes)}) + nodes.interrupt_processing() + + elif behavior == "Mute" or behavior == True: + # mute + should_be_mute_nodes = active_nodes + bypass_nodes + if len(should_be_mute_nodes) > 0: + PromptServer.instance.send_sync("impact-bridge-continue", {"node_id": unique_id, 'mutes': list(should_be_mute_nodes)}) + nodes.interrupt_processing() + + else: + # bypass + should_be_bypass_nodes = active_nodes + mute_nodes + if len(should_be_bypass_nodes) > 0: + PromptServer.instance.send_sync("impact-bridge-continue", {"node_id": unique_id, 'bypasses': list(should_be_bypass_nodes)}) + nodes.interrupt_processing() + + return (value, ) + + +class ImpactExecutionOrderController: + @classmethod + def INPUT_TYPES(cls): + return {"required": { + "signal": (any_typ,), + "value": (any_typ,), + }} + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + RETURN_TYPES = (any_typ, any_typ) + RETURN_NAMES = ("signal", "value") + + def doit(self, signal, value): + return signal, value + + +original_handle_execution = execution.PromptExecutor.handle_execution_error + + +def handle_execution_error(**kwargs): + execution.PromptExecutor.handle_execution_error(**kwargs) + diff --git a/ComfyUI-Impact-Pack/modules/impact/mmdet_nodes.py b/ComfyUI-Impact-Pack/modules/impact/mmdet_nodes.py new file mode 100644 index 0000000000000000000000000000000000000000..81b206d549641d05706116dc8f0cd863e98346ce --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/mmdet_nodes.py @@ -0,0 +1,219 @@ +import folder_paths +from impact.core import * +import os + +import mmcv +from mmdet.apis import (inference_detector, init_detector) +from mmdet.evaluation import get_classes + + +def load_mmdet(model_path): + model_config = os.path.splitext(model_path)[0] + ".py" + model = init_detector(model_config, model_path, device="cpu") + return model + + +def inference_segm_old(model, image, conf_threshold): + image = image.numpy()[0] * 255 + mmdet_results = inference_detector(model, image) + + bbox_results, segm_results = mmdet_results + label = "A" + + classes = get_classes("coco") + labels = [ + np.full(bbox.shape[0], i, dtype=np.int32) + for i, bbox in enumerate(bbox_results) + ] + n, m = bbox_results[0].shape + if n == 0: + return [[], [], []] + labels = np.concatenate(labels) + bboxes = np.vstack(bbox_results) + segms = mmcv.concat_list(segm_results) + filter_idxs = np.where(bboxes[:, -1] > conf_threshold)[0] + results = [[], [], []] + for i in filter_idxs: + results[0].append(label + "-" + classes[labels[i]]) + results[1].append(bboxes[i]) + results[2].append(segms[i]) + + return results + + +def inference_segm(image, modelname, conf_thres, lab="A"): + image = image.numpy()[0] * 255 + mmdet_results = inference_detector(modelname, image).pred_instances + bboxes = mmdet_results.bboxes.numpy() + segms = mmdet_results.masks.numpy() + scores = mmdet_results.scores.numpy() + + classes = get_classes("coco") + + n, m = bboxes.shape + if n == 0: + return [[], [], [], []] + labels = mmdet_results.labels + filter_inds = np.where(mmdet_results.scores > conf_thres)[0] + results = [[], [], [], []] + for i in filter_inds: + results[0].append(lab + "-" + classes[labels[i]]) + results[1].append(bboxes[i]) + results[2].append(segms[i]) + results[3].append(scores[i]) + + return results + + +def inference_bbox(modelname, image, conf_threshold): + image = image.numpy()[0] * 255 + label = "A" + output = inference_detector(modelname, image).pred_instances + cv2_image = np.array(image) + cv2_image = cv2_image[:, :, ::-1].copy() + cv2_gray = cv2.cvtColor(cv2_image, cv2.COLOR_BGR2GRAY) + + segms = [] + for x0, y0, x1, y1 in output.bboxes: + cv2_mask = np.zeros(cv2_gray.shape, np.uint8) + cv2.rectangle(cv2_mask, (int(x0), int(y0)), (int(x1), int(y1)), 255, -1) + cv2_mask_bool = cv2_mask.astype(bool) + segms.append(cv2_mask_bool) + + n, m = output.bboxes.shape + if n == 0: + return [[], [], [], []] + + bboxes = output.bboxes.numpy() + scores = output.scores.numpy() + filter_idxs = np.where(scores > conf_threshold)[0] + results = [[], [], [], []] + for i in filter_idxs: + results[0].append(label) + results[1].append(bboxes[i]) + results[2].append(segms[i]) + results[3].append(scores[i]) + + return results + + +class BBoxDetector: + bbox_model = None + + def __init__(self, bbox_model): + self.bbox_model = bbox_model + + def detect(self, image, threshold, dilation, crop_factor, drop_size=1, detailer_hook=None): + drop_size = max(drop_size, 1) + mmdet_results = inference_bbox(self.bbox_model, image, threshold) + segmasks = create_segmasks(mmdet_results) + + if dilation > 0: + segmasks = dilate_masks(segmasks, dilation) + + items = [] + h = image.shape[1] + w = image.shape[2] + + for x in segmasks: + item_bbox = x[0] + item_mask = x[1] + + y1, x1, y2, x2 = item_bbox + + if x2 - x1 > drop_size and y2 - y1 > drop_size: # minimum dimension must be (2,2) to avoid squeeze issue + crop_region = make_crop_region(w, h, item_bbox, crop_factor) + cropped_image = crop_image(image, crop_region) + cropped_mask = crop_ndarray2(item_mask, crop_region) + confidence = x[2] + # bbox_size = (item_bbox[2]-item_bbox[0],item_bbox[3]-item_bbox[1]) # (w,h) + + item = SEG(cropped_image, cropped_mask, confidence, crop_region, item_bbox, None, None) + + items.append(item) + + shape = image.shape[1], image.shape[2] + return shape, items + + def detect_combined(self, image, threshold, dilation): + mmdet_results = inference_bbox(self.bbox_model, image, threshold) + segmasks = create_segmasks(mmdet_results) + if dilation > 0: + segmasks = dilate_masks(segmasks, dilation) + + return combine_masks(segmasks) + + def setAux(self, x): + pass + + +class SegmDetector(BBoxDetector): + segm_model = None + + def __init__(self, segm_model): + self.segm_model = segm_model + + def detect(self, image, threshold, dilation, crop_factor, drop_size=1, detailer_hook=None): + drop_size = max(drop_size, 1) + mmdet_results = inference_segm(image, self.segm_model, threshold) + segmasks = create_segmasks(mmdet_results) + + if dilation > 0: + segmasks = dilate_masks(segmasks, dilation) + + items = [] + h = image.shape[1] + w = image.shape[2] + for x in segmasks: + item_bbox = x[0] + item_mask = x[1] + + y1, x1, y2, x2 = item_bbox + + if x2 - x1 > drop_size and y2 - y1 > drop_size: # minimum dimension must be (2,2) to avoid squeeze issue + crop_region = make_crop_region(w, h, item_bbox, crop_factor) + cropped_image = crop_image(image, crop_region) + cropped_mask = crop_ndarray2(item_mask, crop_region) + confidence = x[2] + + item = SEG(cropped_image, cropped_mask, confidence, crop_region, item_bbox, None, None) + items.append(item) + + segs = image.shape, items + + if detailer_hook is not None and hasattr(detailer_hook, "post_detection"): + segs = detailer_hook.post_detection(segs) + + return segs + + def detect_combined(self, image, threshold, dilation): + mmdet_results = inference_bbox(self.bbox_model, image, threshold) + segmasks = create_segmasks(mmdet_results) + if dilation > 0: + segmasks = dilate_masks(segmasks, dilation) + + return combine_masks(segmasks) + + def setAux(self, x): + pass + + +class MMDetDetectorProvider: + @classmethod + def INPUT_TYPES(s): + bboxs = ["bbox/"+x for x in folder_paths.get_filename_list("mmdets_bbox")] + segms = ["segm/"+x for x in folder_paths.get_filename_list("mmdets_segm")] + return {"required": {"model_name": (bboxs + segms, )}} + RETURN_TYPES = ("BBOX_DETECTOR", "SEGM_DETECTOR") + FUNCTION = "load_mmdet" + + CATEGORY = "ImpactPack" + + def load_mmdet(self, model_name): + mmdet_path = folder_paths.get_full_path("mmdets", model_name) + model = load_mmdet(mmdet_path) + + if model_name.startswith("bbox"): + return BBoxDetector(model), NO_SEGM_DETECTOR() + else: + return NO_BBOX_DETECTOR(), model \ No newline at end of file diff --git a/ComfyUI-Impact-Pack/modules/impact/onnx.py b/ComfyUI-Impact-Pack/modules/impact/onnx.py new file mode 100644 index 0000000000000000000000000000000000000000..ad0055aef7ad55ad53b5cace1940476f4e2dff50 --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/onnx.py @@ -0,0 +1,38 @@ +import impact.additional_dependencies +from impact.utils import * + +impact.additional_dependencies.ensure_onnx_package() + +try: + import onnxruntime + + def onnx_inference(image, onnx_model): + # prepare image + pil = tensor2pil(image) + image = np.ascontiguousarray(pil) + image = image[:, :, ::-1] # to BGR image + image = image.astype(np.float32) + image -= [103.939, 116.779, 123.68] # 'caffe' mode image preprocessing + + # do detection + onnx_model = onnxruntime.InferenceSession(onnx_model, providers=["CPUExecutionProvider"]) + outputs = onnx_model.run( + [s_i.name for s_i in onnx_model.get_outputs()], + {onnx_model.get_inputs()[0].name: np.expand_dims(image, axis=0)}, + ) + + labels = [op for op in outputs if op.dtype == "int32"][0] + scores = [op for op in outputs if isinstance(op[0][0], np.float32)][0] + boxes = [op for op in outputs if isinstance(op[0][0], np.ndarray)][0] + + # filter-out useless item + idx = np.where(labels[0] == -1)[0][0] + + labels = labels[0][:idx] + scores = scores[0][:idx] + boxes = boxes[0][:idx].astype(np.uint32) + + return labels, scores, boxes +except Exception as e: + print("[ERROR] ComfyUI-Impact-Pack: 'onnxruntime' package doesn't support 'python 3.11', yet.") + print(f"\t{e}") diff --git a/ComfyUI-Impact-Pack/modules/impact/pipe.py b/ComfyUI-Impact-Pack/modules/impact/pipe.py new file mode 100644 index 0000000000000000000000000000000000000000..b98139202ca24125fd5b2c5d14778c7c6445da9d --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/pipe.py @@ -0,0 +1,441 @@ +import folder_paths +import impact.wildcards +from impact.utils import any_typ + + +class ToDetailerPipe: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "model": ("MODEL",), + "clip": ("CLIP",), + "vae": ("VAE",), + "positive": ("CONDITIONING",), + "negative": ("CONDITIONING",), + "bbox_detector": ("BBOX_DETECTOR", ), + "wildcard": ("STRING", {"multiline": True, "dynamicPrompts": False}), + "Select to add LoRA": (["Select the LoRA to add to the text"] + folder_paths.get_filename_list("loras"),), + "Select to add Wildcard": (["Select the Wildcard to add to the text"], ), + }, + "optional": { + "sam_model_opt": ("SAM_MODEL",), + "segm_detector_opt": ("SEGM_DETECTOR",), + "detailer_hook": ("DETAILER_HOOK",), + }} + + RETURN_TYPES = ("DETAILER_PIPE", ) + RETURN_NAMES = ("detailer_pipe", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Pipe" + + def doit(self, *args, **kwargs): + pipe = (kwargs['model'], kwargs['clip'], kwargs['vae'], kwargs['positive'], kwargs['negative'], kwargs['wildcard'], kwargs['bbox_detector'], + kwargs.get('segm_detector_opt', None), kwargs.get('sam_model_opt', None), kwargs.get('detailer_hook', None), + kwargs.get('refiner_model', None), kwargs.get('refiner_clip', None), + kwargs.get('refiner_positive', None), kwargs.get('refiner_negative', None)) + return (pipe, ) + + +class ToDetailerPipeSDXL(ToDetailerPipe): + @classmethod + def INPUT_TYPES(s): + return {"required": { + "model": ("MODEL",), + "clip": ("CLIP",), + "vae": ("VAE",), + "positive": ("CONDITIONING",), + "negative": ("CONDITIONING",), + "refiner_model": ("MODEL",), + "refiner_clip": ("CLIP",), + "refiner_positive": ("CONDITIONING",), + "refiner_negative": ("CONDITIONING",), + "bbox_detector": ("BBOX_DETECTOR", ), + "wildcard": ("STRING", {"multiline": True, "dynamicPrompts": False}), + "Select to add LoRA": (["Select the LoRA to add to the text"] + folder_paths.get_filename_list("loras"),), + "Select to add Wildcard": (["Select the Wildcard to add to the text"],), + }, + "optional": { + "sam_model_opt": ("SAM_MODEL",), + "segm_detector_opt": ("SEGM_DETECTOR",), + "detailer_hook": ("DETAILER_HOOK",), + }} + + +class FromDetailerPipe: + @classmethod + def INPUT_TYPES(s): + return {"required": {"detailer_pipe": ("DETAILER_PIPE",), }, } + + RETURN_TYPES = ("MODEL", "CLIP", "VAE", "CONDITIONING", "CONDITIONING", "BBOX_DETECTOR", "SAM_MODEL", "SEGM_DETECTOR", "DETAILER_HOOK") + RETURN_NAMES = ("model", "clip", "vae", "positive", "negative", "bbox_detector", "sam_model_opt", "segm_detector_opt", "detailer_hook") + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Pipe" + + def doit(self, detailer_pipe): + model, clip, vae, positive, negative, wildcard, bbox_detector, segm_detector_opt, sam_model_opt, detailer_hook, _, _, _, _ = detailer_pipe + return model, clip, vae, positive, negative, bbox_detector, sam_model_opt, segm_detector_opt, detailer_hook + + +class FromDetailerPipe_v2: + @classmethod + def INPUT_TYPES(s): + return {"required": {"detailer_pipe": ("DETAILER_PIPE",), }, } + + RETURN_TYPES = ("DETAILER_PIPE", "MODEL", "CLIP", "VAE", "CONDITIONING", "CONDITIONING", "BBOX_DETECTOR", "SAM_MODEL", "SEGM_DETECTOR", "DETAILER_HOOK") + RETURN_NAMES = ("detailer_pipe", "model", "clip", "vae", "positive", "negative", "bbox_detector", "sam_model_opt", "segm_detector_opt", "detailer_hook") + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Pipe" + + def doit(self, detailer_pipe): + model, clip, vae, positive, negative, wildcard, bbox_detector, segm_detector_opt, sam_model_opt, detailer_hook, _, _, _, _ = detailer_pipe + return detailer_pipe, model, clip, vae, positive, negative, bbox_detector, sam_model_opt, segm_detector_opt, detailer_hook + + +class FromDetailerPipe_SDXL: + @classmethod + def INPUT_TYPES(s): + return {"required": {"detailer_pipe": ("DETAILER_PIPE",), }, } + + RETURN_TYPES = ("DETAILER_PIPE", "MODEL", "CLIP", "VAE", "CONDITIONING", "CONDITIONING", "BBOX_DETECTOR", "SAM_MODEL", "SEGM_DETECTOR", "DETAILER_HOOK", "MODEL", "CLIP", "CONDITIONING", "CONDITIONING") + RETURN_NAMES = ("detailer_pipe", "model", "clip", "vae", "positive", "negative", "bbox_detector", "sam_model_opt", "segm_detector_opt", "detailer_hook", "refiner_model", "refiner_clip", "refiner_positive", "refiner_negative") + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Pipe" + + def doit(self, detailer_pipe): + model, clip, vae, positive, negative, wildcard, bbox_detector, segm_detector_opt, sam_model_opt, detailer_hook, refiner_model, refiner_clip, refiner_positive, refiner_negative = detailer_pipe + return detailer_pipe, model, clip, vae, positive, negative, bbox_detector, sam_model_opt, segm_detector_opt, detailer_hook, refiner_model, refiner_clip, refiner_positive, refiner_negative + + +class AnyPipeToBasic: + @classmethod + def INPUT_TYPES(s): + return { + "required": {"any_pipe": (any_typ,)}, + } + + RETURN_TYPES = ("BASIC_PIPE", ) + RETURN_NAMES = ("basic_pipe", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Pipe" + + def doit(self, any_pipe): + return (any_pipe[:5], ) + + +class ToBasicPipe: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "model": ("MODEL",), + "clip": ("CLIP",), + "vae": ("VAE",), + "positive": ("CONDITIONING",), + "negative": ("CONDITIONING",), + }, + } + + RETURN_TYPES = ("BASIC_PIPE", ) + RETURN_NAMES = ("basic_pipe", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Pipe" + + def doit(self, model, clip, vae, positive, negative): + pipe = (model, clip, vae, positive, negative) + return (pipe, ) + + +class FromBasicPipe: + @classmethod + def INPUT_TYPES(s): + return {"required": {"basic_pipe": ("BASIC_PIPE",), }, } + + RETURN_TYPES = ("MODEL", "CLIP", "VAE", "CONDITIONING", "CONDITIONING") + RETURN_NAMES = ("model", "clip", "vae", "positive", "negative") + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Pipe" + + def doit(self, basic_pipe): + model, clip, vae, positive, negative = basic_pipe + return model, clip, vae, positive, negative + + +class FromBasicPipe_v2: + @classmethod + def INPUT_TYPES(s): + return {"required": {"basic_pipe": ("BASIC_PIPE",), }, } + + RETURN_TYPES = ("BASIC_PIPE", "MODEL", "CLIP", "VAE", "CONDITIONING", "CONDITIONING") + RETURN_NAMES = ("basic_pipe", "model", "clip", "vae", "positive", "negative") + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Pipe" + + def doit(self, basic_pipe): + model, clip, vae, positive, negative = basic_pipe + return basic_pipe, model, clip, vae, positive, negative + + +class BasicPipeToDetailerPipe: + @classmethod + def INPUT_TYPES(s): + return {"required": {"basic_pipe": ("BASIC_PIPE",), + "bbox_detector": ("BBOX_DETECTOR", ), + "wildcard": ("STRING", {"multiline": True, "dynamicPrompts": False}), + "Select to add LoRA": (["Select the LoRA to add to the text"] + folder_paths.get_filename_list("loras"),), + "Select to add Wildcard": (["Select the Wildcard to add to the text"],), + }, + "optional": { + "sam_model_opt": ("SAM_MODEL", ), + "segm_detector_opt": ("SEGM_DETECTOR",), + "detailer_hook": ("DETAILER_HOOK",), + }, + } + + RETURN_TYPES = ("DETAILER_PIPE", ) + RETURN_NAMES = ("detailer_pipe", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Pipe" + + def doit(self, *args, **kwargs): + basic_pipe = kwargs['basic_pipe'] + bbox_detector = kwargs['bbox_detector'] + wildcard = kwargs['wildcard'] + sam_model_opt = kwargs.get('sam_model_opt', None) + segm_detector_opt = kwargs.get('segm_detector_opt', None) + detailer_hook = kwargs.get('detailer_hook', None) + + model, clip, vae, positive, negative = basic_pipe + pipe = model, clip, vae, positive, negative, wildcard, bbox_detector, segm_detector_opt, sam_model_opt, detailer_hook, None, None, None, None + return (pipe, ) + + +class BasicPipeToDetailerPipeSDXL: + @classmethod + def INPUT_TYPES(s): + return {"required": {"base_basic_pipe": ("BASIC_PIPE",), + "refiner_basic_pipe": ("BASIC_PIPE",), + "bbox_detector": ("BBOX_DETECTOR", ), + "wildcard": ("STRING", {"multiline": True, "dynamicPrompts": False}), + "Select to add LoRA": (["Select the LoRA to add to the text"] + folder_paths.get_filename_list("loras"),), + "Select to add Wildcard": (["Select the Wildcard to add to the text"],), + }, + "optional": { + "sam_model_opt": ("SAM_MODEL", ), + "segm_detector_opt": ("SEGM_DETECTOR",), + "detailer_hook": ("DETAILER_HOOK",), + }, + } + + RETURN_TYPES = ("DETAILER_PIPE", ) + RETURN_NAMES = ("detailer_pipe", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Pipe" + + def doit(self, *args, **kwargs): + base_basic_pipe = kwargs['base_basic_pipe'] + refiner_basic_pipe = kwargs['refiner_basic_pipe'] + bbox_detector = kwargs['bbox_detector'] + wildcard = kwargs['wildcard'] + sam_model_opt = kwargs.get('sam_model_opt', None) + segm_detector_opt = kwargs.get('segm_detector_opt', None) + detailer_hook = kwargs.get('detailer_hook', None) + + model, clip, vae, positive, negative = base_basic_pipe + refiner_model, refiner_clip, refiner_vae, refiner_positive, refiner_negative = refiner_basic_pipe + pipe = model, clip, vae, positive, negative, wildcard, bbox_detector, segm_detector_opt, sam_model_opt, detailer_hook, refiner_model, refiner_clip, refiner_positive, refiner_negative + return (pipe, ) + + +class DetailerPipeToBasicPipe: + @classmethod + def INPUT_TYPES(s): + return {"required": {"detailer_pipe": ("DETAILER_PIPE",), }} + + RETURN_TYPES = ("BASIC_PIPE", "BASIC_PIPE") + RETURN_NAMES = ("base_basic_pipe", "refiner_basic_pipe") + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Pipe" + + def doit(self, detailer_pipe): + model, clip, vae, positive, negative, _, _, _, _, _, refiner_model, refiner_clip, refiner_positive, refiner_negative = detailer_pipe + pipe = model, clip, vae, positive, negative + refiner_pipe = refiner_model, refiner_clip, vae, refiner_positive, refiner_negative + return (pipe, refiner_pipe) + + +class EditBasicPipe: + @classmethod + def INPUT_TYPES(s): + return { + "required": {"basic_pipe": ("BASIC_PIPE",), }, + "optional": { + "model": ("MODEL",), + "clip": ("CLIP",), + "vae": ("VAE",), + "positive": ("CONDITIONING",), + "negative": ("CONDITIONING",), + }, + } + + RETURN_TYPES = ("BASIC_PIPE", ) + RETURN_NAMES = ("basic_pipe", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Pipe" + + def doit(self, basic_pipe, model=None, clip=None, vae=None, positive=None, negative=None): + res_model, res_clip, res_vae, res_positive, res_negative = basic_pipe + + if model is not None: + res_model = model + + if clip is not None: + res_clip = clip + + if vae is not None: + res_vae = vae + + if positive is not None: + res_positive = positive + + if negative is not None: + res_negative = negative + + pipe = res_model, res_clip, res_vae, res_positive, res_negative + + return (pipe, ) + + +class EditDetailerPipe: + @classmethod + def INPUT_TYPES(s): + return { + "required": { + "detailer_pipe": ("DETAILER_PIPE",), + "wildcard": ("STRING", {"multiline": True, "dynamicPrompts": False}), + "Select to add LoRA": (["Select the LoRA to add to the text"] + folder_paths.get_filename_list("loras"),), + "Select to add Wildcard": (["Select the Wildcard to add to the text"],), + }, + "optional": { + "model": ("MODEL",), + "clip": ("CLIP",), + "vae": ("VAE",), + "positive": ("CONDITIONING",), + "negative": ("CONDITIONING",), + "bbox_detector": ("BBOX_DETECTOR",), + "sam_model": ("SAM_MODEL",), + "segm_detector": ("SEGM_DETECTOR",), + "detailer_hook": ("DETAILER_HOOK",), + }, + } + + RETURN_TYPES = ("DETAILER_PIPE",) + RETURN_NAMES = ("detailer_pipe",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Pipe" + + def doit(self, *args, **kwargs): + detailer_pipe = kwargs['detailer_pipe'] + wildcard = kwargs['wildcard'] + model = kwargs.get('model', None) + clip = kwargs.get('clip', None) + vae = kwargs.get('vae', None) + positive = kwargs.get('positive', None) + negative = kwargs.get('negative', None) + bbox_detector = kwargs.get('bbox_detector', None) + sam_model = kwargs.get('sam_model', None) + segm_detector = kwargs.get('segm_detector', None) + detailer_hook = kwargs.get('detailer_hook', None) + refiner_model = kwargs.get('refiner_model', None) + refiner_clip = kwargs.get('refiner_clip', None) + refiner_positive = kwargs.get('refiner_positive', None) + refiner_negative = kwargs.get('refiner_negative', None) + + res_model, res_clip, res_vae, res_positive, res_negative, res_wildcard, res_bbox_detector, res_segm_detector, res_sam_model, res_detailer_hook, res_refiner_model, res_refiner_clip, res_refiner_positive, res_refiner_negative = detailer_pipe + + if model is not None: + res_model = model + + if clip is not None: + res_clip = clip + + if vae is not None: + res_vae = vae + + if positive is not None: + res_positive = positive + + if negative is not None: + res_negative = negative + + if bbox_detector is not None: + res_bbox_detector = bbox_detector + + if segm_detector is not None: + res_segm_detector = segm_detector + + if wildcard != "": + res_wildcard = wildcard + + if sam_model is not None: + res_sam_model = sam_model + + if detailer_hook is not None: + res_detailer_hook = detailer_hook + + if refiner_model is not None: + res_refiner_model = refiner_model + + if refiner_clip is not None: + res_refiner_clip = refiner_clip + + if refiner_positive is not None: + res_refiner_positive = refiner_positive + + if refiner_negative is not None: + res_refiner_negative = refiner_negative + + pipe = (res_model, res_clip, res_vae, res_positive, res_negative, res_wildcard, + res_bbox_detector, res_segm_detector, res_sam_model, res_detailer_hook, + res_refiner_model, res_refiner_clip, res_refiner_positive, res_refiner_negative) + + return (pipe, ) + + +class EditDetailerPipeSDXL(EditDetailerPipe): + @classmethod + def INPUT_TYPES(s): + return { + "required": { + "detailer_pipe": ("DETAILER_PIPE",), + "wildcard": ("STRING", {"multiline": True, "dynamicPrompts": False}), + "Select to add LoRA": (["Select the LoRA to add to the text"] + folder_paths.get_filename_list("loras"),), + "Select to add Wildcard": (["Select the Wildcard to add to the text"],), + }, + "optional": { + "model": ("MODEL",), + "clip": ("CLIP",), + "vae": ("VAE",), + "positive": ("CONDITIONING",), + "negative": ("CONDITIONING",), + "refiner_model": ("MODEL",), + "refiner_clip": ("CLIP",), + "refiner_positive": ("CONDITIONING",), + "refiner_negative": ("CONDITIONING",), + "bbox_detector": ("BBOX_DETECTOR",), + "sam_model": ("SAM_MODEL",), + "segm_detector": ("SEGM_DETECTOR",), + "detailer_hook": ("DETAILER_HOOK",), + }, + } diff --git a/ComfyUI-Impact-Pack/modules/impact/sample_error_enhancer.py b/ComfyUI-Impact-Pack/modules/impact/sample_error_enhancer.py new file mode 100644 index 0000000000000000000000000000000000000000..484c12998fad0e8e60c050420ce63227af44ba2f --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/sample_error_enhancer.py @@ -0,0 +1,25 @@ +import comfy.sample +import traceback + +original_sample = comfy.sample.sample + + +def informative_sample(*args, **kwargs): + try: + return original_sample(*args, **kwargs) # This code helps interpret error messages that occur within exceptions but does not have any impact on other operations. + except RuntimeError as e: + is_model_mix_issue = False + try: + if 'mat1 and mat2 shapes cannot be multiplied' in e.args[0]: + if 'torch.nn.functional.linear' in traceback.format_exc().strip().split('\n')[-3]: + is_model_mix_issue = True + except: + pass + + if is_model_mix_issue: + raise RuntimeError("\n\n#### It seems that models and clips are mixed and interconnected between SDXL Base, SDXL Refiner, SD1.x, and SD2.x. Please verify. ####\n\n") + else: + raise e + + +comfy.sample.sample = informative_sample diff --git a/ComfyUI-Impact-Pack/modules/impact/segs_nodes.py b/ComfyUI-Impact-Pack/modules/impact/segs_nodes.py new file mode 100644 index 0000000000000000000000000000000000000000..233c3ed67b15dcf57dd6c93adf95c228573896a8 --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/segs_nodes.py @@ -0,0 +1,1918 @@ +import os +import sys + +import impact.impact_server +from nodes import MAX_RESOLUTION + +from impact.utils import * +from . import core +from .core import SEG +import impact.utils as utils +from . import defs +from . import segs_upscaler +from comfy.cli_args import args +import math + + +try: + from comfy_extras import nodes_differential_diffusion +except Exception: + print(f"\n#############################################\n[Impact Pack] ComfyUI is an outdated version.\n#############################################\n") + raise Exception("[Impact Pack] ComfyUI is an outdated version.") + + +class SEGSDetailer: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "image": ("IMAGE", ), + "segs": ("SEGS", ), + "guide_size": ("FLOAT", {"default": 512, "min": 64, "max": MAX_RESOLUTION, "step": 8}), + "guide_size_for": ("BOOLEAN", {"default": True, "label_on": "bbox", "label_off": "crop_region"}), + "max_size": ("FLOAT", {"default": 768, "min": 64, "max": MAX_RESOLUTION, "step": 8}), + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000}), + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS,), + "scheduler": (core.SCHEDULERS,), + "denoise": ("FLOAT", {"default": 0.5, "min": 0.0001, "max": 1.0, "step": 0.01}), + "noise_mask": ("BOOLEAN", {"default": True, "label_on": "enabled", "label_off": "disabled"}), + "force_inpaint": ("BOOLEAN", {"default": True, "label_on": "enabled", "label_off": "disabled"}), + "basic_pipe": ("BASIC_PIPE",), + "refiner_ratio": ("FLOAT", {"default": 0.2, "min": 0.0, "max": 1.0}), + "batch_size": ("INT", {"default": 1, "min": 1, "max": 100}), + + "cycle": ("INT", {"default": 1, "min": 1, "max": 10, "step": 1}), + }, + "optional": { + "refiner_basic_pipe_opt": ("BASIC_PIPE",), + "inpaint_model": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + "noise_mask_feather": ("INT", {"default": 20, "min": 0, "max": 100, "step": 1}), + "scheduler_func_opt": ("SCHEDULER_FUNC",), + } + } + + RETURN_TYPES = ("SEGS", "IMAGE") + RETURN_NAMES = ("segs", "cnet_images") + OUTPUT_IS_LIST = (False, True) + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detailer" + + @staticmethod + def do_detail(image, segs, guide_size, guide_size_for, max_size, seed, steps, cfg, sampler_name, scheduler, + denoise, noise_mask, force_inpaint, basic_pipe, refiner_ratio=None, batch_size=1, cycle=1, + refiner_basic_pipe_opt=None, inpaint_model=False, noise_mask_feather=0, scheduler_func_opt=None): + + model, clip, vae, positive, negative = basic_pipe + if refiner_basic_pipe_opt is None: + refiner_model, refiner_clip, refiner_positive, refiner_negative = None, None, None, None + else: + refiner_model, refiner_clip, _, refiner_positive, refiner_negative = refiner_basic_pipe_opt + + segs = core.segs_scale_match(segs, image.shape) + + new_segs = [] + cnet_pil_list = [] + + if noise_mask_feather > 0 and 'denoise_mask_function' not in model.model_options: + model = nodes_differential_diffusion.DifferentialDiffusion().apply(model)[0] + + for i in range(batch_size): + seed += 1 + for seg in segs[1]: + cropped_image = seg.cropped_image if seg.cropped_image is not None \ + else crop_ndarray4(image.numpy(), seg.crop_region) + cropped_image = to_tensor(cropped_image) + + is_mask_all_zeros = (seg.cropped_mask == 0).all().item() + if is_mask_all_zeros: + print(f"Detailer: segment skip [empty mask]") + new_segs.append(seg) + continue + + if noise_mask: + cropped_mask = seg.cropped_mask + else: + cropped_mask = None + + cropped_positive = [ + [condition, { + k: core.crop_condition_mask(v, image, seg.crop_region) if k == "mask" else v + for k, v in details.items() + }] + for condition, details in positive + ] + + cropped_negative = [ + [condition, { + k: core.crop_condition_mask(v, image, seg.crop_region) if k == "mask" else v + for k, v in details.items() + }] + for condition, details in negative + ] + + enhanced_image, cnet_pils = core.enhance_detail(cropped_image, model, clip, vae, guide_size, guide_size_for, max_size, + seg.bbox, seed, steps, cfg, sampler_name, scheduler, + cropped_positive, cropped_negative, denoise, cropped_mask, force_inpaint, + refiner_ratio=refiner_ratio, refiner_model=refiner_model, + refiner_clip=refiner_clip, refiner_positive=refiner_positive, refiner_negative=refiner_negative, + control_net_wrapper=seg.control_net_wrapper, cycle=cycle, + inpaint_model=inpaint_model, noise_mask_feather=noise_mask_feather, scheduler_func=scheduler_func_opt) + + if cnet_pils is not None: + cnet_pil_list.extend(cnet_pils) + + if enhanced_image is None: + new_cropped_image = cropped_image + else: + new_cropped_image = enhanced_image + + new_seg = SEG(to_numpy(new_cropped_image), seg.cropped_mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, None) + new_segs.append(new_seg) + + return (segs[0], new_segs), cnet_pil_list + + def doit(self, image, segs, guide_size, guide_size_for, max_size, seed, steps, cfg, sampler_name, scheduler, + denoise, noise_mask, force_inpaint, basic_pipe, refiner_ratio=None, batch_size=1, cycle=1, + refiner_basic_pipe_opt=None, inpaint_model=False, noise_mask_feather=0, scheduler_func_opt=None): + + if len(image) > 1: + raise Exception('[Impact Pack] ERROR: SEGSDetailer does not allow image batches.\nPlease refer to https://github.com/ltdrdata/ComfyUI-extension-tutorials/blob/Main/ComfyUI-Impact-Pack/tutorial/batching-detailer.md for more information.') + + segs, cnet_pil_list = SEGSDetailer.do_detail(image, segs, guide_size, guide_size_for, max_size, seed, steps, cfg, sampler_name, + scheduler, denoise, noise_mask, force_inpaint, basic_pipe, refiner_ratio, batch_size, cycle=cycle, + refiner_basic_pipe_opt=refiner_basic_pipe_opt, + inpaint_model=inpaint_model, noise_mask_feather=noise_mask_feather, scheduler_func_opt=scheduler_func_opt) + + # set fallback image + if len(cnet_pil_list) == 0: + cnet_pil_list = [empty_pil_tensor()] + + return segs, cnet_pil_list + + +class SEGSPaste: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "image": ("IMAGE", ), + "segs": ("SEGS", ), + "feather": ("INT", {"default": 5, "min": 0, "max": 100, "step": 1}), + "alpha": ("INT", {"default": 255, "min": 0, "max": 255, "step": 1}), + }, + "optional": {"ref_image_opt": ("IMAGE", ), } + } + + RETURN_TYPES = ("IMAGE", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Detailer" + + @staticmethod + def doit(image, segs, feather, alpha=255, ref_image_opt=None): + + segs = core.segs_scale_match(segs, image.shape) + + result = None + for i, single_image in enumerate(image): + image_i = single_image.unsqueeze(0).clone() + + for seg in segs[1]: + ref_image = None + if ref_image_opt is None and seg.cropped_image is not None: + cropped_image = seg.cropped_image + if isinstance(cropped_image, np.ndarray): + cropped_image = torch.from_numpy(cropped_image) + ref_image = cropped_image[i].unsqueeze(0) + elif ref_image_opt is not None: + ref_tensor = ref_image_opt[i].unsqueeze(0) + ref_image = crop_image(ref_tensor, seg.crop_region) + if ref_image is not None: + if seg.cropped_mask.ndim == 3 and len(seg.cropped_mask) == len(image): + mask = seg.cropped_mask[i] + elif seg.cropped_mask.ndim == 3 and len(seg.cropped_mask) > 1: + print(f"[Impact Pack] WARN: SEGSPaste - The number of the mask batch({len(seg.cropped_mask)}) and the image batch({len(image)}) are different. Combine the mask frames and apply.") + combined_mask = (seg.cropped_mask[0] * 255).to(torch.uint8) + + for frame_mask in seg.cropped_mask[1:]: + combined_mask |= (frame_mask * 255).to(torch.uint8) + + combined_mask = (combined_mask/255.0).to(torch.float32) + mask = utils.to_binary_mask(combined_mask, 0.1) + else: # ndim == 2 + mask = seg.cropped_mask + + mask = tensor_gaussian_blur_mask(mask, feather) * (alpha/255) + x, y, *_ = seg.crop_region + + # ensure same device + mask = mask.to(image_i.device) + ref_image = ref_image.to(image_i.device) + + tensor_paste(image_i, ref_image, (x, y), mask) + + if result is None: + result = image_i + else: + result = torch.concat((result, image_i), dim=0) + + if not args.highvram and not args.gpu_only: + result = result.cpu() + + return (result, ) + + +class SEGSPreviewCNet: + def __init__(self): + self.output_dir = folder_paths.get_temp_directory() + self.type = "temp" + + @classmethod + def INPUT_TYPES(s): + return {"required": {"segs": ("SEGS", ),}, } + + RETURN_TYPES = ("IMAGE", ) + OUTPUT_IS_LIST = (True, ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + OUTPUT_NODE = True + + def doit(self, segs): + full_output_folder, filename, counter, subfolder, filename_prefix = \ + folder_paths.get_save_image_path("impact_seg_preview", self.output_dir, segs[0][1], segs[0][0]) + + results = list() + result_image_list = [] + + for seg in segs[1]: + file = f"{filename}_{counter:05}_.webp" + + if seg.control_net_wrapper is not None and seg.control_net_wrapper.control_image is not None: + cnet_image = seg.control_net_wrapper.control_image + result_image_list.append(cnet_image) + else: + cnet_image = empty_pil_tensor(64, 64) + + cnet_pil = utils.tensor2pil(cnet_image) + cnet_pil.save(os.path.join(full_output_folder, file)) + + results.append({ + "filename": file, + "subfolder": subfolder, + "type": self.type + }) + + counter += 1 + + return {"ui": {"images": results}, "result": (result_image_list,)} + + +class SEGSPreview: + def __init__(self): + self.output_dir = folder_paths.get_temp_directory() + self.type = "temp" + + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS", ), + "alpha_mode": ("BOOLEAN", {"default": True, "label_on": "enable", "label_off": "disable"}), + "min_alpha": ("FLOAT", {"default": 0.2, "min": 0.0, "max": 1.0, "step": 0.01}), + }, + "optional": { + "fallback_image_opt": ("IMAGE", ), + } + } + + RETURN_TYPES = ("IMAGE", ) + OUTPUT_IS_LIST = (True, ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + OUTPUT_NODE = True + + def doit(self, segs, alpha_mode=True, min_alpha=0.0, fallback_image_opt=None): + full_output_folder, filename, counter, subfolder, filename_prefix = \ + folder_paths.get_save_image_path("impact_seg_preview", self.output_dir, segs[0][1], segs[0][0]) + + results = list() + result_image_list = [] + + if fallback_image_opt is not None: + segs = core.segs_scale_match(segs, fallback_image_opt.shape) + + if min_alpha != 0: + min_alpha = int(255 * min_alpha) + + if len(segs[1]) > 0: + if segs[1][0].cropped_image is not None: + batch_count = len(segs[1][0].cropped_image) + elif fallback_image_opt is not None: + batch_count = len(fallback_image_opt) + else: + return {"ui": {"images": results}} + + for seg in segs[1]: + result_image_batch = None + cached_mask = None + + def get_combined_mask(): + nonlocal cached_mask + + if cached_mask is not None: + return cached_mask + else: + if isinstance(seg.cropped_mask, np.ndarray): + masks = torch.tensor(seg.cropped_mask) + else: + masks = seg.cropped_mask + + cached_mask = (masks[0] * 255).to(torch.uint8) + for x in masks[1:]: + cached_mask |= (x * 255).to(torch.uint8) + cached_mask = (cached_mask/255.0).to(torch.float32) + cached_mask = utils.to_binary_mask(cached_mask, 0.1) + cached_mask = cached_mask.numpy() + + return cached_mask + + def stack_image(image, mask=None): + nonlocal result_image_batch + + if isinstance(image, np.ndarray): + image = torch.from_numpy(image) + + if mask is not None: + image *= torch.tensor(mask)[None, ..., None] + + if result_image_batch is None: + result_image_batch = image + else: + result_image_batch = torch.concat((result_image_batch, image), dim=0) + + for i in range(batch_count): + cropped_image = None + + if seg.cropped_image is not None: + cropped_image = seg.cropped_image[i, None] + elif fallback_image_opt is not None: + # take from original image + ref_image = fallback_image_opt[i].unsqueeze(0) + cropped_image = crop_image(ref_image, seg.crop_region) + + if cropped_image is not None: + if isinstance(cropped_image, np.ndarray): + cropped_image = torch.from_numpy(cropped_image) + + cropped_image = cropped_image.clone() + cropped_pil = to_pil(cropped_image) + + if alpha_mode: + if isinstance(seg.cropped_mask, np.ndarray): + cropped_mask = seg.cropped_mask + else: + if seg.cropped_image is not None and len(seg.cropped_image) != len(seg.cropped_mask): + cropped_mask = get_combined_mask() + else: + cropped_mask = seg.cropped_mask[i].numpy() + + mask_array = (cropped_mask * 255).astype(np.uint8) + + if min_alpha != 0: + mask_array[mask_array < min_alpha] = min_alpha + + mask_pil = Image.fromarray(mask_array, mode='L').resize(cropped_pil.size) + cropped_pil.putalpha(mask_pil) + stack_image(cropped_image, cropped_mask) + else: + stack_image(cropped_image) + + file = f"{filename}_{counter:05}_.webp" + cropped_pil.save(os.path.join(full_output_folder, file)) + results.append({ + "filename": file, + "subfolder": subfolder, + "type": self.type + }) + + counter += 1 + + if result_image_batch is not None: + result_image_list.append(result_image_batch) + + return {"ui": {"images": results}, "result": (result_image_list,) } + + +class SEGSLabelFilter: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS", ), + "preset": (['all'] + defs.detection_labels, ), + "labels": ("STRING", {"multiline": True, "placeholder": "List the types of segments to be allowed, separated by commas"}), + }, + } + + RETURN_TYPES = ("SEGS", "SEGS",) + RETURN_NAMES = ("filtered_SEGS", "remained_SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + @staticmethod + def filter(segs, labels): + labels = set([label.strip() for label in labels]) + + if 'all' in labels: + return (segs, (segs[0], []), ) + else: + res_segs = [] + remained_segs = [] + + for x in segs[1]: + if x.label in labels: + res_segs.append(x) + elif 'eyes' in labels and x.label in ['left_eye', 'right_eye']: + res_segs.append(x) + elif 'eyebrows' in labels and x.label in ['left_eyebrow', 'right_eyebrow']: + res_segs.append(x) + elif 'pupils' in labels and x.label in ['left_pupil', 'right_pupil']: + res_segs.append(x) + else: + remained_segs.append(x) + + return ((segs[0], res_segs), (segs[0], remained_segs), ) + + def doit(self, segs, preset, labels): + labels = labels.split(',') + return SEGSLabelFilter.filter(segs, labels) + + +class SEGSLabelAssign: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS", ), + "labels": ("STRING", {"multiline": True, "placeholder": "List the label to be assigned in order of segs, separated by commas"}), + }, + } + + RETURN_TYPES = ("SEGS",) + RETURN_NAMES = ("SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + @staticmethod + def assign(segs, labels): + labels = [label.strip() for label in labels] + + if len(labels) != len(segs[1]): + print(f'Warning (SEGSLabelAssign): length of labels ({len(labels)}) != length of segs ({len(segs[1])})') + + labeled_segs = [] + + idx = 0 + for x in segs[1]: + if len(labels) > idx: + x = x._replace(label=labels[idx]) + labeled_segs.append(x) + idx += 1 + + return ((segs[0], labeled_segs), ) + + def doit(self, segs, labels): + labels = labels.split(',') + return SEGSLabelAssign.assign(segs, labels) + + +class SEGSOrderedFilter: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS", ), + "target": (["area(=w*h)", "width", "height", "x1", "y1", "x2", "y2", "confidence"],), + "order": ("BOOLEAN", {"default": True, "label_on": "descending", "label_off": "ascending"}), + "take_start": ("INT", {"default": 0, "min": 0, "max": sys.maxsize, "step": 1}), + "take_count": ("INT", {"default": 1, "min": 0, "max": sys.maxsize, "step": 1}), + }, + } + + RETURN_TYPES = ("SEGS", "SEGS",) + RETURN_NAMES = ("filtered_SEGS", "remained_SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, segs, target, order, take_start, take_count): + segs_with_order = [] + + for seg in segs[1]: + x1 = seg.crop_region[0] + y1 = seg.crop_region[1] + x2 = seg.crop_region[2] + y2 = seg.crop_region[3] + + if target == "area(=w*h)": + value = (y2 - y1) * (x2 - x1) + elif target == "width": + value = x2 - x1 + elif target == "height": + value = y2 - y1 + elif target == "x1": + value = x1 + elif target == "x2": + value = x2 + elif target == "y1": + value = y1 + elif target == "y2": + value = y2 + elif target == "confidence": + value = seg.confidence + else: + raise Exception(f"[Impact Pack] SEGSOrderedFilter - Unexpected target '{target}'") + + segs_with_order.append((value, seg)) + + if order: + sorted_list = sorted(segs_with_order, key=lambda x: x[0], reverse=True) + else: + sorted_list = sorted(segs_with_order, key=lambda x: x[0], reverse=False) + + result_list = [] + remained_list = [] + + for i, item in enumerate(sorted_list): + if take_start <= i < take_start + take_count: + result_list.append(item[1]) + else: + remained_list.append(item[1]) + + return (segs[0], result_list), (segs[0], remained_list), + + +class SEGSRangeFilter: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS", ), + "target": (["area(=w*h)", "width", "height", "x1", "y1", "x2", "y2", "length_percent", "confidence(0-100)"],), + "mode": ("BOOLEAN", {"default": True, "label_on": "inside", "label_off": "outside"}), + "min_value": ("INT", {"default": 0, "min": 0, "max": sys.maxsize, "step": 1}), + "max_value": ("INT", {"default": 67108864, "min": 0, "max": sys.maxsize, "step": 1}), + }, + } + + RETURN_TYPES = ("SEGS", "SEGS",) + RETURN_NAMES = ("filtered_SEGS", "remained_SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, segs, target, mode, min_value, max_value): + new_segs = [] + remained_segs = [] + + for seg in segs[1]: + x1 = seg.crop_region[0] + y1 = seg.crop_region[1] + x2 = seg.crop_region[2] + y2 = seg.crop_region[3] + + if target == "area(=w*h)": + value = (y2 - y1) * (x2 - x1) + elif target == "length_percent": + h = y2 - y1 + w = x2 - x1 + value = max(h/w, w/h)*100 + print(f"value={value}") + elif target == "width": + value = x2 - x1 + elif target == "height": + value = y2 - y1 + elif target == "x1": + value = x1 + elif target == "x2": + value = x2 + elif target == "y1": + value = y1 + elif target == "y2": + value = y2 + elif target == "confidence(0-100)": + value = seg.confidence*100 + else: + raise Exception(f"[Impact Pack] SEGSRangeFilter - Unexpected target '{target}'") + + if mode and min_value <= value <= max_value: + print(f"[in] value={value} / {mode}, {min_value}, {max_value}") + new_segs.append(seg) + elif not mode and (value < min_value or value > max_value): + print(f"[out] value={value} / {mode}, {min_value}, {max_value}") + new_segs.append(seg) + else: + remained_segs.append(seg) + print(f"[filter] value={value} / {mode}, {min_value}, {max_value}") + + return (segs[0], new_segs), (segs[0], remained_segs), + + +class SEGSToImageList: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS", ), + }, + "optional": { + "fallback_image_opt": ("IMAGE", ), + } + } + + RETURN_TYPES = ("IMAGE",) + OUTPUT_IS_LIST = (True,) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, segs, fallback_image_opt=None): + results = list() + + if fallback_image_opt is not None: + segs = core.segs_scale_match(segs, fallback_image_opt.shape) + + for seg in segs[1]: + if seg.cropped_image is not None: + cropped_image = to_tensor(seg.cropped_image) + elif fallback_image_opt is not None: + # take from original image + cropped_image = to_tensor(crop_image(fallback_image_opt, seg.crop_region)) + else: + cropped_image = empty_pil_tensor() + + results.append(cropped_image) + + if len(results) == 0: + results.append(empty_pil_tensor()) + + return (results,) + + +class SEGSToMaskList: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS", ), + }, + } + + RETURN_TYPES = ("MASK",) + OUTPUT_IS_LIST = (True,) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, segs): + masks = core.segs_to_masklist(segs) + if len(masks) == 0: + empty_mask = torch.zeros(segs[0], dtype=torch.float32, device="cpu") + masks = [empty_mask] + masks = [utils.make_3d_mask(mask) for mask in masks] + return (masks,) + + +class SEGSToMaskBatch: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS", ), + }, + } + + RETURN_TYPES = ("MASK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, segs): + masks = core.segs_to_masklist(segs) + masks = [utils.make_3d_mask(mask) for mask in masks] + mask_batch = torch.concat(masks) + return (mask_batch,) + + +class SEGSMerge: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS", ), + }, + } + + RETURN_TYPES = ("SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + DESCRIPTION = "SEGS contains multiple SEGs. SEGS Merge integrates several SEGs into a single merged SEG. The label is changed to `merged` and the confidence becomes the minimum confidence. The applied controlnet and cropped_image are removed." + + def doit(self, segs): + crop_left = sys.maxsize + crop_right = 0 + crop_top = sys.maxsize + crop_bottom = 0 + + bbox_left = sys.maxsize + bbox_right = 0 + bbox_top = sys.maxsize + bbox_bottom = 0 + + min_confidence = 1.0 + + for seg in segs[1]: + cx1 = seg.crop_region[0] + cy1 = seg.crop_region[1] + cx2 = seg.crop_region[2] + cy2 = seg.crop_region[3] + + bx1 = seg.bbox[0] + by1 = seg.bbox[1] + bx2 = seg.bbox[2] + by2 = seg.bbox[3] + + crop_left = min(crop_left, cx1) + crop_top = min(crop_top, cy1) + crop_right = max(crop_right, cx2) + crop_bottom = max(crop_bottom, cy2) + + bbox_left = min(bbox_left, bx1) + bbox_top = min(bbox_top, by1) + bbox_right = max(bbox_right, bx2) + bbox_bottom = max(bbox_bottom, by2) + + min_confidence = min(min_confidence, seg.confidence) + + combined_mask = core.segs_to_combined_mask(segs) + cropped_mask = combined_mask[crop_top:crop_bottom, crop_left:crop_right] + cropped_mask = cropped_mask.unsqueeze(0) + + crop_region = [crop_left, crop_top, crop_right, crop_bottom] + bbox = [bbox_left, bbox_top, bbox_right, bbox_bottom] + + seg = SEG(None, cropped_mask, min_confidence, crop_region, bbox, 'merged', None) + return ((segs[0], [seg]),) + + +class SEGSConcat: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs1": ("SEGS", ), + }, + } + + RETURN_TYPES = ("SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, **kwargs): + dim = None + res = None + + for k, v in list(kwargs.items()): + if v[0] == (0, 0) or len(v[1]) == 0: + continue + + if dim is None: + dim = v[0] + res = v[1] + else: + if v[0] == dim: + res = res + v[1] + else: + print(f"ERROR: source shape of 'segs1'{dim} and '{k}'{v[0]} are different. '{k}' will be ignored") + + if dim is None: + empty_segs = ((0, 0), []) + return (empty_segs, ) + else: + return ((dim, res), ) + + +class Count_Elts_in_SEGS: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS", ), + }, + } + + RETURN_TYPES = ("INT",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, segs): + return (len(segs[1]), ) + + +class DecomposeSEGS: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS", ), + }, + } + + RETURN_TYPES = ("SEGS_HEADER", "SEG_ELT",) + OUTPUT_IS_LIST = (False, True, ) + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, segs): + return segs + + +class AssembleSEGS: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "seg_header": ("SEGS_HEADER", ), + "seg_elt": ("SEG_ELT", ), + }, + } + + INPUT_IS_LIST = True + + RETURN_TYPES = ("SEGS", ) + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, seg_header, seg_elt): + return ((seg_header[0], seg_elt), ) + + +class From_SEG_ELT: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "seg_elt": ("SEG_ELT", ), + }, + } + + RETURN_TYPES = ("SEG_ELT", "IMAGE", "MASK", "SEG_ELT_crop_region", "SEG_ELT_bbox", "SEG_ELT_control_net_wrapper", "FLOAT", "STRING") + RETURN_NAMES = ("seg_elt", "cropped_image", "cropped_mask", "crop_region", "bbox", "control_net_wrapper", "confidence", "label") + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, seg_elt): + cropped_image = to_tensor(seg_elt.cropped_image) if seg_elt.cropped_image is not None else None + return (seg_elt, cropped_image, to_tensor(seg_elt.cropped_mask), seg_elt.crop_region, seg_elt.bbox, seg_elt.control_net_wrapper, seg_elt.confidence, seg_elt.label,) + + +class From_SEG_ELT_bbox: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "bbox": ("SEG_ELT_bbox", ), + }, + } + + RETURN_TYPES = ("INT", "INT", "INT", "INT") + RETURN_NAMES = ("left", "top", "right", "bottom") + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, bbox): + return bbox + + +class From_SEG_ELT_crop_region: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "crop_region": ("SEG_ELT_crop_region", ), + }, + } + + RETURN_TYPES = ("INT", "INT", "INT", "INT") + RETURN_NAMES = ("left", "top", "right", "bottom") + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, crop_region): + return crop_region + + +class Edit_SEG_ELT: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "seg_elt": ("SEG_ELT", ), + }, + "optional": { + "cropped_image_opt": ("IMAGE", ), + "cropped_mask_opt": ("MASK", ), + "crop_region_opt": ("SEG_ELT_crop_region", ), + "bbox_opt": ("SEG_ELT_bbox", ), + "control_net_wrapper_opt": ("SEG_ELT_control_net_wrapper", ), + "confidence_opt": ("FLOAT", {"min": 0, "max": 1.0, "step": 0.1, "forceInput": True}), + "label_opt": ("STRING", {"multiline": False, "forceInput": True}), + } + } + + RETURN_TYPES = ("SEG_ELT", ) + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, seg_elt, cropped_image_opt=None, cropped_mask_opt=None, confidence_opt=None, crop_region_opt=None, + bbox_opt=None, label_opt=None, control_net_wrapper_opt=None): + + cropped_image = seg_elt.cropped_image if cropped_image_opt is None else cropped_image_opt + cropped_mask = seg_elt.cropped_mask if cropped_mask_opt is None else cropped_mask_opt + confidence = seg_elt.confidence if confidence_opt is None else confidence_opt + crop_region = seg_elt.crop_region if crop_region_opt is None else crop_region_opt + bbox = seg_elt.bbox if bbox_opt is None else bbox_opt + label = seg_elt.label if label_opt is None else label_opt + control_net_wrapper = seg_elt.control_net_wrapper if control_net_wrapper_opt is None else control_net_wrapper_opt + + cropped_image = cropped_image.numpy() if cropped_image is not None else None + + if isinstance(cropped_mask, torch.Tensor): + if len(cropped_mask.shape) == 3: + cropped_mask = cropped_mask.squeeze(0) + + cropped_mask = cropped_mask.numpy() + + seg = SEG(cropped_image, cropped_mask, confidence, crop_region, bbox, label, control_net_wrapper) + + return (seg,) + + +class DilateMask: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "mask": ("MASK", ), + "dilation": ("INT", {"default": 10, "min": -512, "max": 512, "step": 1}), + }} + + RETURN_TYPES = ("MASK", ) + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, mask, dilation): + mask = core.dilate_mask(mask.numpy(), dilation) + mask = torch.from_numpy(mask) + mask = utils.make_3d_mask(mask) + return (mask, ) + + +class GaussianBlurMask: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "mask": ("MASK", ), + "kernel_size": ("INT", {"default": 10, "min": 0, "max": 100, "step": 1}), + "sigma": ("FLOAT", {"default": 10.0, "min": 0.1, "max": 100.0, "step": 0.1}), + }} + + RETURN_TYPES = ("MASK", ) + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, mask, kernel_size, sigma): + # Some custom nodes use abnormal 4-dimensional masks in the format of b, c, h, w. In the impact pack, internal 4-dimensional masks are required in the format of b, h, w, c. Therefore, normalization is performed using the normal mask format, which is 3-dimensional, before proceeding with the operation. + mask = make_3d_mask(mask) + mask = torch.unsqueeze(mask, dim=-1) + mask = utils.tensor_gaussian_blur_mask(mask, kernel_size, sigma) + mask = torch.squeeze(mask, dim=-1) + return (mask, ) + + +class DilateMaskInSEGS: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS", ), + "dilation": ("INT", {"default": 10, "min": -512, "max": 512, "step": 1}), + }} + + RETURN_TYPES = ("SEGS", ) + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, segs, dilation): + new_segs = [] + for seg in segs[1]: + mask = core.dilate_mask(seg.cropped_mask, dilation) + seg = SEG(seg.cropped_image, mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, seg.control_net_wrapper) + new_segs.append(seg) + + return ((segs[0], new_segs), ) + + +class GaussianBlurMaskInSEGS: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS", ), + "kernel_size": ("INT", {"default": 10, "min": 0, "max": 100, "step": 1}), + "sigma": ("FLOAT", {"default": 10.0, "min": 0.1, "max": 100.0, "step": 0.1}), + }} + + RETURN_TYPES = ("SEGS", ) + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, segs, kernel_size, sigma): + new_segs = [] + for seg in segs[1]: + mask = utils.tensor_gaussian_blur_mask(seg.cropped_mask, kernel_size, sigma) + mask = torch.squeeze(mask, dim=-1).squeeze(0).numpy() + seg = SEG(seg.cropped_image, mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, seg.control_net_wrapper) + new_segs.append(seg) + + return ((segs[0], new_segs), ) + + +class Dilate_SEG_ELT: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "seg_elt": ("SEG_ELT", ), + "dilation": ("INT", {"default": 10, "min": -512, "max": 512, "step": 1}), + }} + + RETURN_TYPES = ("SEG_ELT", ) + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, seg, dilation): + mask = core.dilate_mask(seg.cropped_mask, dilation) + seg = SEG(seg.cropped_image, mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, seg.control_net_wrapper) + return (seg,) + + +class SEG_ELT_BBOX_ScaleBy: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "seg": ("SEG_ELT", ), + "scale_by": ("FLOAT", {"default": 1.0, "min": 0.01, "max": 8.0, "step": 0.01}), } + } + + RETURN_TYPES = ("SEG_ELT", ) + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + @staticmethod + def fill_zero_outside_bbox(mask, crop_region, bbox): + cx1, cy1, _, _ = crop_region + x1, y1, x2, y2 = bbox + x1, y1, x2, y2 = x1-cx1, y1-cy1, x2-cx1, y2-cy1 + h, w = mask.shape + + x1 = min(w-1, max(0, x1)) + x2 = min(w-1, max(0, x2)) + y1 = min(h-1, max(0, y1)) + y2 = min(h-1, max(0, y2)) + + mask_cropped = mask.copy() + mask_cropped[:, :x1] = 0 # zero fill left side + mask_cropped[:, x2:] = 0 # zero fill right side + mask_cropped[:y1, :] = 0 # zero fill top side + mask_cropped[y2:, :] = 0 # zero fill bottom side + return mask_cropped + + def doit(self, seg, scale_by): + x1, y1, x2, y2 = seg.bbox + w = x2-x1 + h = y2-y1 + + dw = int((w * scale_by - w)/2) + dh = int((h * scale_by - h)/2) + + bbox = (x1-dw, y1-dh, x2+dw, y2+dh) + + cropped_mask = SEG_ELT_BBOX_ScaleBy.fill_zero_outside_bbox(seg.cropped_mask, seg.crop_region, bbox) + seg = SEG(seg.cropped_image, cropped_mask, seg.confidence, seg.crop_region, bbox, seg.label, seg.control_net_wrapper) + return (seg,) + + +class EmptySEGS: + @classmethod + def INPUT_TYPES(s): + return {"required": {}, } + + RETURN_TYPES = ("SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self): + shape = 0, 0 + return ((shape, []),) + + +class SegsToCombinedMask: + @classmethod + def INPUT_TYPES(s): + return {"required": {"segs": ("SEGS",), }} + + RETURN_TYPES = ("MASK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Operation" + + def doit(self, segs): + mask = core.segs_to_combined_mask(segs) + mask = utils.make_3d_mask(mask) + return (mask,) + + +class MediaPipeFaceMeshToSEGS: + @classmethod + def INPUT_TYPES(s): + bool_true_widget = ("BOOLEAN", {"default": True, "label_on": "Enabled", "label_off": "Disabled"}) + bool_false_widget = ("BOOLEAN", {"default": False, "label_on": "Enabled", "label_off": "Disabled"}) + return {"required": { + "image": ("IMAGE",), + "crop_factor": ("FLOAT", {"default": 3.0, "min": 1.0, "max": 100, "step": 0.1}), + "bbox_fill": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + "crop_min_size": ("INT", {"min": 10, "max": MAX_RESOLUTION, "step": 1, "default": 50}), + "drop_size": ("INT", {"min": 1, "max": MAX_RESOLUTION, "step": 1, "default": 1}), + "dilation": ("INT", {"default": 0, "min": -512, "max": 512, "step": 1}), + "face": bool_true_widget, + "mouth": bool_false_widget, + "left_eyebrow": bool_false_widget, + "left_eye": bool_false_widget, + "left_pupil": bool_false_widget, + "right_eyebrow": bool_false_widget, + "right_eye": bool_false_widget, + "right_pupil": bool_false_widget, + }, + # "optional": {"reference_image_opt": ("IMAGE", ), } + } + + RETURN_TYPES = ("SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Operation" + + def doit(self, image, crop_factor, bbox_fill, crop_min_size, drop_size, dilation, face, mouth, left_eyebrow, left_eye, left_pupil, right_eyebrow, right_eye, right_pupil): + # padding is obsolete now + # https://github.com/Fannovel16/comfyui_controlnet_aux/blob/1ec41fceff1ee99596445a0c73392fd91df407dc/utils.py#L33 + # def calc_pad(h_raw, w_raw): + # resolution = normalize_size_base_64(h_raw, w_raw) + # + # def pad64(x): + # return int(np.ceil(float(x) / 64.0) * 64 - x) + # + # k = float(resolution) / float(min(h_raw, w_raw)) + # h_target = int(np.round(float(h_raw) * k)) + # w_target = int(np.round(float(w_raw) * k)) + # + # return pad64(h_target), pad64(w_target) + + # if reference_image_opt is not None: + # if image.shape[1:] != reference_image_opt.shape[1:]: + # scale_by1 = reference_image_opt.shape[1] / image.shape[1] + # scale_by2 = reference_image_opt.shape[2] / image.shape[2] + # scale_by = min(scale_by1, scale_by2) + # + # # padding is obsolete now + # # h_pad, w_pad = calc_pad(reference_image_opt.shape[1], reference_image_opt.shape[2]) + # # if h_pad != 0: + # # # height padded + # # image = image[:, :-h_pad, :, :] + # # elif w_pad != 0: + # # # width padded + # # image = image[:, :, :-w_pad, :] + # + # image = nodes.ImageScaleBy().upscale(image, "bilinear", scale_by)[0] + + result = core.mediapipe_facemesh_to_segs(image, crop_factor, bbox_fill, crop_min_size, drop_size, dilation, face, mouth, left_eyebrow, left_eye, left_pupil, right_eyebrow, right_eye, right_pupil) + return (result, ) + + +class MaskToSEGS: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "mask": ("MASK",), + "combined": ("BOOLEAN", {"default": False, "label_on": "True", "label_off": "False"}), + "crop_factor": ("FLOAT", {"default": 3.0, "min": 1.0, "max": 100, "step": 0.1}), + "bbox_fill": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + "drop_size": ("INT", {"min": 1, "max": MAX_RESOLUTION, "step": 1, "default": 10}), + "contour_fill": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + } + } + + RETURN_TYPES = ("SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Operation" + + @staticmethod + def doit(mask, combined, crop_factor, bbox_fill, drop_size, contour_fill=False): + mask = make_2d_mask(mask) + result = core.mask_to_segs(mask, combined, crop_factor, bbox_fill, drop_size, is_contour=contour_fill) + + return (result, ) + + +class MaskToSEGS_for_AnimateDiff: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "mask": ("MASK",), + "combined": ("BOOLEAN", {"default": False, "label_on": "True", "label_off": "False"}), + "crop_factor": ("FLOAT", {"default": 3.0, "min": 1.0, "max": 100, "step": 0.1}), + "bbox_fill": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + "drop_size": ("INT", {"min": 1, "max": MAX_RESOLUTION, "step": 1, "default": 10}), + "contour_fill": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + } + } + + RETURN_TYPES = ("SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Operation" + + @staticmethod + def doit(mask, combined, crop_factor, bbox_fill, drop_size, contour_fill=False): + if (len(mask.shape) == 4 and mask.shape[1] > 1) or (len(mask.shape) == 3 and mask.shape[0] > 1): + mask = make_3d_mask(mask) + if contour_fill: + print(f"[Impact Pack] MaskToSEGS_for_AnimateDiff: 'contour_fill' is ignored because batch mask 'contour_fill' is not supported.") + result = core.batch_mask_to_segs(mask, combined, crop_factor, bbox_fill, drop_size) + return (result, ) + + mask = make_2d_mask(mask) + segs = core.mask_to_segs(mask, combined, crop_factor, bbox_fill, drop_size, is_contour=contour_fill) + all_masks = SEGSToMaskList().doit(segs)[0] + + result_mask = (all_masks[0] * 255).to(torch.uint8) + for mask in all_masks[1:]: + result_mask |= (mask * 255).to(torch.uint8) + + result_mask = (result_mask/255.0).to(torch.float32) + result_mask = utils.to_binary_mask(result_mask, 0.1)[0] + + return MaskToSEGS.doit(result_mask, False, crop_factor, False, drop_size, contour_fill) + + +class IPAdapterApplySEGS: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS",), + "ipadapter_pipe": ("IPADAPTER_PIPE",), + "weight": ("FLOAT", {"default": 0.7, "min": -1, "max": 3, "step": 0.05}), + "noise": ("FLOAT", {"default": 0.4, "min": 0.0, "max": 1.0, "step": 0.01}), + "weight_type": (["original", "linear", "channel penalty"], {"default": 'channel penalty'}), + "start_at": ("FLOAT", {"default": 0.0, "min": 0.0, "max": 1.0, "step": 0.001}), + "end_at": ("FLOAT", {"default": 0.9, "min": 0.0, "max": 1.0, "step": 0.001}), + "unfold_batch": ("BOOLEAN", {"default": False}), + "faceid_v2": ("BOOLEAN", {"default": False}), + "weight_v2": ("FLOAT", {"default": 1.0, "min": -1, "max": 3, "step": 0.05}), + "context_crop_factor": ("FLOAT", {"default": 1.2, "min": 1.0, "max": 100, "step": 0.1}), + "reference_image": ("IMAGE",), + }, + "optional": { + "combine_embeds": (["concat", "add", "subtract", "average", "norm average"],), + "neg_image": ("IMAGE",), + }, + } + + RETURN_TYPES = ("SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + @staticmethod + def doit(segs, ipadapter_pipe, weight, noise, weight_type, start_at, end_at, unfold_batch, faceid_v2, weight_v2, context_crop_factor, reference_image, combine_embeds="concat", neg_image=None): + + if len(ipadapter_pipe) == 4: + print(f"[Impact Pack] IPAdapterApplySEGS: Installed Inspire Pack is outdated.") + raise Exception("Inspire Pack is outdated.") + + new_segs = [] + + h, w = segs[0] + + if reference_image.shape[2] != w or reference_image.shape[1] != h: + reference_image = tensor_resize(reference_image, w, h) + + for seg in segs[1]: + # The context_crop_region sets how much wider the IPAdapter context will reflect compared to the crop_region, not the bbox + context_crop_region = make_crop_region(w, h, seg.crop_region, context_crop_factor) + cropped_image = crop_image(reference_image, context_crop_region) + + control_net_wrapper = core.IPAdapterWrapper(ipadapter_pipe, weight, noise, weight_type, start_at, end_at, unfold_batch, weight_v2, cropped_image, neg_image=neg_image, prev_control_net=seg.control_net_wrapper, combine_embeds=combine_embeds) + new_seg = SEG(seg.cropped_image, seg.cropped_mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, control_net_wrapper) + new_segs.append(new_seg) + + return ((segs[0], new_segs), ) + + +class ControlNetApplySEGS: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS",), + "control_net": ("CONTROL_NET",), + "strength": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 10.0, "step": 0.01}), + }, + "optional": { + "segs_preprocessor": ("SEGS_PREPROCESSOR",), + "control_image": ("IMAGE",) + } + } + + RETURN_TYPES = ("SEGS",) + FUNCTION = "doit" + + DEPRECATED = True + + CATEGORY = "ImpactPack/Util" + + @staticmethod + def doit(segs, control_net, strength, segs_preprocessor=None, control_image=None): + new_segs = [] + + for seg in segs[1]: + control_net_wrapper = core.ControlNetWrapper(control_net, strength, segs_preprocessor, seg.control_net_wrapper, + original_size=segs[0], crop_region=seg.crop_region, control_image=control_image) + new_seg = SEG(seg.cropped_image, seg.cropped_mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, control_net_wrapper) + new_segs.append(new_seg) + + return ((segs[0], new_segs), ) + + +class ControlNetApplyAdvancedSEGS: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS",), + "control_net": ("CONTROL_NET",), + "strength": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 10.0, "step": 0.01}), + "start_percent": ("FLOAT", {"default": 0.0, "min": 0.0, "max": 1.0, "step": 0.001}), + "end_percent": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.001}) + }, + "optional": { + "segs_preprocessor": ("SEGS_PREPROCESSOR",), + "control_image": ("IMAGE",), + "vae": ("VAE",) + } + } + + RETURN_TYPES = ("SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + @staticmethod + def doit(segs, control_net, strength, start_percent, end_percent, segs_preprocessor=None, control_image=None, vae=None): + new_segs = [] + + for seg in segs[1]: + control_net_wrapper = core.ControlNetAdvancedWrapper(control_net, strength, start_percent, end_percent, segs_preprocessor, + seg.control_net_wrapper, original_size=segs[0], crop_region=seg.crop_region, + control_image=control_image, vae=vae) + new_seg = SEG(seg.cropped_image, seg.cropped_mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, control_net_wrapper) + new_segs.append(new_seg) + + return ((segs[0], new_segs), ) + + +class ControlNetClearSEGS: + @classmethod + def INPUT_TYPES(s): + return {"required": {"segs": ("SEGS",), }, } + + RETURN_TYPES = ("SEGS",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + @staticmethod + def doit(segs): + new_segs = [] + + for seg in segs[1]: + new_seg = SEG(seg.cropped_image, seg.cropped_mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, None) + new_segs.append(new_seg) + + return ((segs[0], new_segs), ) + + +class SEGSSwitch: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "select": ("INT", {"default": 1, "min": 1, "max": 99999, "step": 1}), + "segs1": ("SEGS",), + }, + } + + RETURN_TYPES = ("SEGS", ) + + OUTPUT_NODE = True + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, *args, **kwargs): + input_name = f"segs{int(kwargs['select'])}" + + if input_name in kwargs: + return (kwargs[input_name],) + else: + print(f"SEGSSwitch: invalid select index ('segs1' is selected)") + return (kwargs['segs1'],) + + +class SEGSPicker: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "picks": ("STRING", {"multiline": True, "dynamicPrompts": False, "pysssss.autocomplete": False}), + "segs": ("SEGS",), + }, + "optional": { + "fallback_image_opt": ("IMAGE", ), + }, + "hidden": {"unique_id": "UNIQUE_ID"}, + } + + RETURN_TYPES = ("SEGS", ) + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + @staticmethod + def doit(picks, segs, fallback_image_opt=None, unique_id=None): + if fallback_image_opt is not None: + segs = core.segs_scale_match(segs, fallback_image_opt.shape) + + # generate candidates image + cands = [] + for seg in segs[1]: + if seg.cropped_image is not None: + cropped_image = seg.cropped_image + elif fallback_image_opt is not None: + # take from original image + cropped_image = crop_image(fallback_image_opt, seg.crop_region) + else: + cropped_image = empty_pil_tensor() + + mask_array = seg.cropped_mask.copy() + mask_array[mask_array < 0.3] = 0.3 + mask_array = mask_array[None, ..., None] + cropped_image = cropped_image * mask_array + + cands.append(cropped_image) + + impact.impact_server.segs_picker_map[unique_id] = cands + + # pass only selected + pick_ids = set() + + for pick in picks.split(","): + try: + pick_ids.add(int(pick)-1) + except Exception: + pass + + new_segs = [] + for i in pick_ids: + if 0 <= i < len(segs[1]): + new_segs.append(segs[1][i]) + + return ((segs[0], new_segs),) + + +class DefaultImageForSEGS: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "segs": ("SEGS", ), + "image": ("IMAGE", ), + "override": ("BOOLEAN", {"default": True}), + }} + + RETURN_TYPES = ("SEGS", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + @staticmethod + def doit(segs, image, override): + results = [] + + segs = core.segs_scale_match(segs, image.shape) + + if len(segs[1]) > 0: + if segs[1][0].cropped_image is not None: + batch_count = len(segs[1][0].cropped_image) + else: + batch_count = len(image) + + for seg in segs[1]: + if seg.cropped_image is not None and not override: + cropped_image = seg.cropped_image + else: + cropped_image = None + for i in range(0, batch_count): + # take from original image + ref_image = image[i].unsqueeze(0) + cropped_image2 = crop_image(ref_image, seg.crop_region) + + if cropped_image is None: + cropped_image = cropped_image2 + else: + cropped_image = torch.cat((cropped_image, cropped_image2), dim=0) + + new_seg = SEG(cropped_image, seg.cropped_mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, seg.control_net_wrapper) + results.append(new_seg) + + return ((segs[0], results), ) + else: + return (segs, ) + + +class RemoveImageFromSEGS: + @classmethod + def INPUT_TYPES(s): + return {"required": {"segs": ("SEGS", ), }} + + RETURN_TYPES = ("SEGS", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + @staticmethod + def doit(segs): + results = [] + + if len(segs[1]) > 0: + for seg in segs[1]: + new_seg = SEG(None, seg.cropped_mask, seg.confidence, seg.crop_region, seg.bbox, seg.label, seg.control_net_wrapper) + results.append(new_seg) + + return ((segs[0], results), ) + else: + return (segs, ) + + +class MakeTileSEGS: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "images": ("IMAGE", ), + "bbox_size": ("INT", {"default": 512, "min": 64, "max": 4096, "step": 8}), + "crop_factor": ("FLOAT", {"default": 3.0, "min": 1.0, "max": 10, "step": 0.01}), + "min_overlap": ("INT", {"default": 5, "min": 0, "max": 512, "step": 1}), + "filter_segs_dilation": ("INT", {"default": 20, "min": -255, "max": 255, "step": 1}), + "mask_irregularity": ("FLOAT", {"default": 0, "min": 0, "max": 1.0, "step": 0.01}), + "irregular_mask_mode": (["Reuse fast", "Reuse quality", "All random fast", "All random quality"],) + }, + "optional": { + "filter_in_segs_opt": ("SEGS", ), + "filter_out_segs_opt": ("SEGS", ), + } + } + + RETURN_TYPES = ("SEGS",) + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/__for_testing" + + @staticmethod + def doit(images, bbox_size, crop_factor, min_overlap, filter_segs_dilation, mask_irregularity=0, irregular_mask_mode="Reuse fast", filter_in_segs_opt=None, filter_out_segs_opt=None): + if bbox_size <= 2*min_overlap: + new_min_overlap = bbox_size / 2 + print(f"[MakeTileSEGS] min_overlap should be greater than bbox_size. (value changed: {min_overlap} => {new_min_overlap})") + min_overlap = new_min_overlap + + _, ih, iw, _ = images.size() + + mask_cache = None + mask_quality = 512 + if mask_irregularity > 0: + if irregular_mask_mode == "Reuse fast": + mask_quality = 128 + mask_cache = np.zeros((128, 128)).astype(np.float32) + core.random_mask(mask_cache, (0, 0, 128, 128), factor=mask_irregularity, size=mask_quality) + elif irregular_mask_mode == "Reuse quality": + mask_quality = 512 + mask_cache = np.zeros((512, 512)).astype(np.float32) + core.random_mask(mask_cache, (0, 0, 512, 512), factor=mask_irregularity, size=mask_quality) + elif irregular_mask_mode == "All random fast": + mask_quality = 512 + + # compensate overlap/bbox_size for irregular mask + if mask_irregularity > 0: + compensate = max(6, int(mask_quality * mask_irregularity / 4)) + min_overlap += compensate + bbox_size += compensate*2 + + # create exclusion mask + if filter_out_segs_opt is not None: + exclusion_mask = core.segs_to_combined_mask(filter_out_segs_opt) + exclusion_mask = utils.make_3d_mask(exclusion_mask) + exclusion_mask = utils.resize_mask(exclusion_mask, (ih, iw)) + exclusion_mask = dilate_mask(exclusion_mask.cpu().numpy(), filter_segs_dilation) + else: + exclusion_mask = None + + if filter_in_segs_opt is not None: + and_mask = core.segs_to_combined_mask(filter_in_segs_opt) + and_mask = utils.make_3d_mask(and_mask) + and_mask = utils.resize_mask(and_mask, (ih, iw)) + and_mask = dilate_mask(and_mask.cpu().numpy(), filter_segs_dilation) + + a, b = core.mask_to_segs(and_mask, True, 1.0, False, 0) + if len(b) == 0: + return ((a, b),) + + start_x, start_y, c, d = b[0].crop_region + w = c - start_x + h = d - start_y + else: + start_x = 0 + start_y = 0 + h, w = ih, iw + and_mask = None + + # calculate tile factors + if bbox_size > h or bbox_size > w: + new_bbox_size = min(bbox_size, min(w, h)) + print(f"[MaskTileSEGS] bbox_size is greater than resolution (value changed: {bbox_size} => {new_bbox_size}") + bbox_size = new_bbox_size + + n_horizontal = math.ceil(w / (bbox_size - min_overlap)) + n_vertical = math.ceil(h / (bbox_size - min_overlap)) + + w_overlap_sum = (bbox_size * n_horizontal) - w + if w_overlap_sum < 0: + n_horizontal += 1 + w_overlap_sum = (bbox_size * n_horizontal) - w + + w_overlap_size = 0 if n_horizontal == 1 else int(w_overlap_sum/(n_horizontal-1)) + + h_overlap_sum = (bbox_size * n_vertical) - h + if h_overlap_sum < 0: + n_vertical += 1 + h_overlap_sum = (bbox_size * n_vertical) - h + + h_overlap_size = 0 if n_vertical == 1 else int(h_overlap_sum/(n_vertical-1)) + + new_segs = [] + + if w_overlap_size == bbox_size: + n_horizontal = 1 + + if h_overlap_size == bbox_size: + n_vertical = 1 + + y = start_y + for j in range(0, n_vertical): + x = start_x + for i in range(0, n_horizontal): + x1 = x + y1 = y + + if x+bbox_size < iw-1: + x2 = x+bbox_size + else: + x2 = iw + x1 = iw-bbox_size + + if y+bbox_size < ih-1: + y2 = y+bbox_size + else: + y2 = ih + y1 = ih-bbox_size + + bbox = x1, y1, x2, y2 + crop_region = make_crop_region(iw, ih, bbox, crop_factor) + cx1, cy1, cx2, cy2 = crop_region + + mask = np.zeros((cy2 - cy1, cx2 - cx1)).astype(np.float32) + + rel_left = x1 - cx1 + rel_top = y1 - cy1 + rel_right = x2 - cx1 + rel_bot = y2 - cy1 + + if mask_irregularity > 0: + if mask_cache is not None: + core.adaptive_mask_paste(mask, mask_cache, (rel_left, rel_top, rel_right, rel_bot)) + else: + core.random_mask(mask, (rel_left, rel_top, rel_right, rel_bot), factor=mask_irregularity, size=mask_quality) + + # corner filling + if rel_left == 0: + pad = int((x2 - x1) / 8) + mask[rel_top:rel_bot, :pad] = 1.0 + + if rel_top == 0: + pad = int((y2 - y1) / 8) + mask[:pad, rel_left:rel_right] = 1.0 + + if rel_right == mask.shape[1]: + pad = int((x2 - x1) / 8) + mask[rel_top:rel_bot, -pad:] = 1.0 + + if rel_bot == mask.shape[0]: + pad = int((y2 - y1) / 8) + mask[-pad:, rel_left:rel_right] = 1.0 + else: + mask[rel_top:rel_bot, rel_left:rel_right] = 1.0 + + mask = torch.tensor(mask) + + if exclusion_mask is not None: + exclusion_mask_cropped = exclusion_mask[cy1:cy2, cx1:cx2] + mask[exclusion_mask_cropped != 0] = 0.0 + + if and_mask is not None: + and_mask_cropped = and_mask[cy1:cy2, cx1:cx2] + mask[and_mask_cropped == 0] = 0.0 + + is_mask_zero = torch.all(mask == 0.0).item() + + if not is_mask_zero: + item = SEG(None, mask.numpy(), 1.0, crop_region, bbox, "", None) + new_segs.append(item) + + x += bbox_size - w_overlap_size + y += bbox_size - h_overlap_size + + res = (ih, iw), new_segs # segs + return (res,) + + +class SEGSUpscaler: + @classmethod + def INPUT_TYPES(s): + resampling_methods = ["lanczos", "nearest", "bilinear", "bicubic"] + + return {"required": { + "image": ("IMAGE",), + "segs": ("SEGS",), + "model": ("MODEL",), + "clip": ("CLIP",), + "vae": ("VAE",), + "rescale_factor": ("FLOAT", {"default": 2, "min": 0.01, "max": 100.0, "step": 0.01}), + "resampling_method": (resampling_methods,), + "supersample": (["true", "false"],), + "rounding_modulus": ("INT", {"default": 8, "min": 8, "max": 1024, "step": 8}), + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000}), + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS,), + "scheduler": (core.SCHEDULERS,), + "positive": ("CONDITIONING",), + "negative": ("CONDITIONING",), + "denoise": ("FLOAT", {"default": 0.5, "min": 0.0001, "max": 1.0, "step": 0.01}), + "feather": ("INT", {"default": 5, "min": 0, "max": 100, "step": 1}), + "inpaint_model": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + "noise_mask_feather": ("INT", {"default": 20, "min": 0, "max": 100, "step": 1}), + }, + "optional": { + "upscale_model_opt": ("UPSCALE_MODEL",), + "upscaler_hook_opt": ("UPSCALER_HOOK",), + "scheduler_func_opt": ("SCHEDULER_FUNC",), + } + } + + RETURN_TYPES = ("IMAGE",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Upscale" + + @staticmethod + def doit(image, segs, model, clip, vae, rescale_factor, resampling_method, supersample, rounding_modulus, + seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise, feather, inpaint_model, noise_mask_feather, + upscale_model_opt=None, upscaler_hook_opt=None, scheduler_func_opt=None): + + new_image = segs_upscaler.upscaler(image, upscale_model_opt, rescale_factor, resampling_method, supersample, rounding_modulus) + + segs = core.segs_scale_match(segs, new_image.shape) + + ordered_segs = segs[1] + + for i, seg in enumerate(ordered_segs): + cropped_image = crop_ndarray4(new_image.numpy(), seg.crop_region) + cropped_image = to_tensor(cropped_image) + mask = to_tensor(seg.cropped_mask) + mask = tensor_gaussian_blur_mask(mask, feather) + + is_mask_all_zeros = (seg.cropped_mask == 0).all().item() + if is_mask_all_zeros: + print(f"SEGSUpscaler: segment skip [empty mask]") + continue + + cropped_mask = seg.cropped_mask + + seg_seed = seed + i + + enhanced_image = segs_upscaler.img2img_segs(cropped_image, model, clip, vae, seg_seed, steps, cfg, sampler_name, scheduler, + positive, negative, denoise, + noise_mask=cropped_mask, control_net_wrapper=seg.control_net_wrapper, + inpaint_model=inpaint_model, noise_mask_feather=noise_mask_feather, scheduler_func_opt=scheduler_func_opt) + if not (enhanced_image is None): + new_image = new_image.cpu() + enhanced_image = enhanced_image.cpu() + left = seg.crop_region[0] + top = seg.crop_region[1] + tensor_paste(new_image, enhanced_image, (left, top), mask) + + if upscaler_hook_opt is not None: + new_image = upscaler_hook_opt.post_paste(new_image) + + enhanced_img = tensor_convert_rgb(new_image) + + return (enhanced_img,) + + +class SEGSUpscalerPipe: + @classmethod + def INPUT_TYPES(s): + resampling_methods = ["lanczos", "nearest", "bilinear", "bicubic"] + + return {"required": { + "image": ("IMAGE",), + "segs": ("SEGS",), + "basic_pipe": ("BASIC_PIPE",), + "rescale_factor": ("FLOAT", {"default": 2, "min": 0.01, "max": 100.0, "step": 0.01}), + "resampling_method": (resampling_methods,), + "supersample": (["true", "false"],), + "rounding_modulus": ("INT", {"default": 8, "min": 8, "max": 1024, "step": 8}), + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000}), + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS,), + "scheduler": (core.SCHEDULERS,), + "denoise": ("FLOAT", {"default": 0.5, "min": 0.0001, "max": 1.0, "step": 0.01}), + "feather": ("INT", {"default": 5, "min": 0, "max": 100, "step": 1}), + "inpaint_model": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + "noise_mask_feather": ("INT", {"default": 20, "min": 0, "max": 100, "step": 1}), + }, + "optional": { + "upscale_model_opt": ("UPSCALE_MODEL",), + "upscaler_hook_opt": ("UPSCALER_HOOK",), + "scheduler_func_opt": ("SCHEDULER_FUNC",), + } + } + + RETURN_TYPES = ("IMAGE",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Upscale" + + @staticmethod + def doit(image, segs, basic_pipe, rescale_factor, resampling_method, supersample, rounding_modulus, + seed, steps, cfg, sampler_name, scheduler, denoise, feather, inpaint_model, noise_mask_feather, + upscale_model_opt=None, upscaler_hook_opt=None, scheduler_func_opt=None): + + model, clip, vae, positive, negative = basic_pipe + + return SEGSUpscaler.doit(image, segs, model, clip, vae, rescale_factor, resampling_method, supersample, rounding_modulus, + seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise, feather, inpaint_model, noise_mask_feather, + upscale_model_opt=upscale_model_opt, upscaler_hook_opt=upscaler_hook_opt, scheduler_func_opt=scheduler_func_opt) diff --git a/ComfyUI-Impact-Pack/modules/impact/segs_upscaler.py b/ComfyUI-Impact-Pack/modules/impact/segs_upscaler.py new file mode 100644 index 0000000000000000000000000000000000000000..8cfa8fb0e7bb0fc7cca1b680e935951438059c18 --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/segs_upscaler.py @@ -0,0 +1,132 @@ +from impact.utils import * +from impact import impact_sampling +from comfy import model_management +from comfy.cli_args import args +import nodes + +try: + from comfy_extras import nodes_differential_diffusion +except Exception: + print(f"[Impact Pack] ComfyUI is an outdated version. The DifferentialDiffusion feature will be disabled.") + + +# Implementation based on `https://github.com/lingondricka2/Upscaler-Detailer` + +# code from comfyroll ---> +# https://github.com/Suzie1/ComfyUI_Comfyroll_CustomNodes/blob/main/nodes/functions_upscale.py + +def upscale_with_model(upscale_model, image): + device = model_management.get_torch_device() + upscale_model.to(device) + in_img = image.movedim(-1, -3).to(device) + free_memory = model_management.get_free_memory(device) + + tile = 512 + overlap = 32 + + oom = True + while oom: + try: + steps = in_img.shape[0] * comfy.utils.get_tiled_scale_steps(in_img.shape[3], in_img.shape[2], tile_x=tile, tile_y=tile, overlap=overlap) + pbar = comfy.utils.ProgressBar(steps) + s = comfy.utils.tiled_scale(in_img, lambda a: upscale_model(a), tile_x=tile, tile_y=tile, overlap=overlap, upscale_amount=upscale_model.scale, pbar=pbar) + oom = False + except model_management.OOM_EXCEPTION as e: + tile //= 2 + if tile < 128: + raise e + + s = torch.clamp(s.movedim(-3, -1), min=0, max=1.0) + return s + + +def apply_resize_image(image: Image.Image, original_width, original_height, rounding_modulus, mode='scale', supersample='true', factor: int = 2, width: int = 1024, height: int = 1024, + resample='bicubic'): + # Calculate the new width and height based on the given mode and parameters + if mode == 'rescale': + new_width, new_height = int(original_width * factor), int(original_height * factor) + else: + m = rounding_modulus + original_ratio = original_height / original_width + height = int(width * original_ratio) + + new_width = width if width % m == 0 else width + (m - width % m) + new_height = height if height % m == 0 else height + (m - height % m) + + # Define a dictionary of resampling filters + resample_filters = {'nearest': 0, 'bilinear': 2, 'bicubic': 3, 'lanczos': 1} + + # Apply supersample + if supersample == 'true': + image = image.resize((new_width * 8, new_height * 8), resample=Image.Resampling(resample_filters[resample])) + + # Resize the image using the given resampling filter + resized_image = image.resize((new_width, new_height), resample=Image.Resampling(resample_filters[resample])) + + return resized_image + + +def upscaler(image, upscale_model, rescale_factor, resampling_method, supersample, rounding_modulus): + if upscale_model is not None: + up_image = upscale_with_model(upscale_model, image) + else: + up_image = image + + pil_img = tensor2pil(image) + original_width, original_height = pil_img.size + scaled_image = pil2tensor(apply_resize_image(tensor2pil(up_image), original_width, original_height, rounding_modulus, 'rescale', + supersample, rescale_factor, 1024, resampling_method)) + return scaled_image + +# <--- + + +def img2img_segs(image, model, clip, vae, seed, steps, cfg, sampler_name, scheduler, + positive, negative, denoise, noise_mask, control_net_wrapper=None, + inpaint_model=False, noise_mask_feather=0, scheduler_func_opt=None): + + original_image_size = image.shape[1:3] + + # Match to original image size + if original_image_size[0] % 8 > 0 or original_image_size[1] % 8 > 0: + scale = 8/min(original_image_size[0], original_image_size[1]) + 1 + w = int(original_image_size[1] * scale) + h = int(original_image_size[0] * scale) + image = tensor_resize(image, w, h) + + if noise_mask is not None: + noise_mask = tensor_gaussian_blur_mask(noise_mask, noise_mask_feather) + noise_mask = noise_mask.squeeze(3) + + if noise_mask_feather > 0 and 'denoise_mask_function' not in model.model_options: + model = nodes_differential_diffusion.DifferentialDiffusion().apply(model)[0] + + if control_net_wrapper is not None: + positive, negative, _ = control_net_wrapper.apply(positive, negative, image, noise_mask) + + # prepare mask + if noise_mask is not None and inpaint_model: + positive, negative, latent_image = nodes.InpaintModelConditioning().encode(positive, negative, image, vae, noise_mask) + else: + latent_image = to_latent_image(image, vae) + if noise_mask is not None: + latent_image['noise_mask'] = noise_mask + + refined_latent = latent_image + + # ksampler + refined_latent = impact_sampling.ksampler_wrapper(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, refined_latent, denoise, scheduler_func=scheduler_func_opt) + + # non-latent downscale - latent downscale cause bad quality + refined_image = vae.decode(refined_latent['samples']) + + # prevent mixing of device + refined_image = refined_image.cpu() + + # Match to original image size + if refined_image.shape[1:3] != original_image_size: + refined_image = tensor_resize(refined_image, original_image_size[1], original_image_size[0]) + + # don't convert to latent - latent break image + # preserving pil is much better + return refined_image diff --git a/ComfyUI-Impact-Pack/modules/impact/special_samplers.py b/ComfyUI-Impact-Pack/modules/impact/special_samplers.py new file mode 100644 index 0000000000000000000000000000000000000000..8f722ddc8e75ca5b275e26d989627d251fdc3838 --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/special_samplers.py @@ -0,0 +1,683 @@ +import math +import impact.core as core +from comfy_extras.nodes_custom_sampler import Noise_RandomNoise +from impact.utils import * +from nodes import MAX_RESOLUTION +import nodes +from impact.impact_sampling import KSamplerWrapper, KSamplerAdvancedWrapper, separated_sample, impact_sample +import comfy + +class TiledKSamplerProvider: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff, "tooltip": "Random seed to use for generating CPU noise for sampling."}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000, "tooltip": "total sampling steps"}), + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0, "tooltip": "classifier free guidance value"}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS, {"tooltip": "sampler"}), + "scheduler": (comfy.samplers.KSampler.SCHEDULERS, {"tooltip": "noise schedule"}), + "denoise": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01, "tooltip": "The amount of noise to remove. This amount is the noise added at the start, and the higher it is, the more the input latent will be modified before being returned."}), + "tile_width": ("INT", {"default": 512, "min": 320, "max": MAX_RESOLUTION, "step": 64, "tooltip": "Sets the width of the tile to be used in TiledKSampler."}), + "tile_height": ("INT", {"default": 512, "min": 320, "max": MAX_RESOLUTION, "step": 64, "tooltip": "Sets the height of the tile to be used in TiledKSampler."}), + "tiling_strategy": (["random", "padded", 'simple'], {"tooltip": "Sets the tiling strategy for TiledKSampler."} ), + "basic_pipe": ("BASIC_PIPE", {"tooltip": "basic_pipe input for sampling"}) + }} + + OUTPUT_TOOLTIPS = ("sampler wrapper. (Can be used when generating a regional_prompt.)", ) + + RETURN_TYPES = ("KSAMPLER",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Sampler" + + @staticmethod + def doit(seed, steps, cfg, sampler_name, scheduler, denoise, + tile_width, tile_height, tiling_strategy, basic_pipe): + model, _, _, positive, negative = basic_pipe + sampler = core.TiledKSamplerWrapper(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise, + tile_width, tile_height, tiling_strategy) + return (sampler, ) + + +class KSamplerProvider: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff, "tooltip": "Random seed to use for generating CPU noise for sampling."}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000, "tooltip": "total sampling steps"}), + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0, "tooltip": "classifier free guidance value"}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS, {"tooltip": "sampler"}), + "scheduler": (core.SCHEDULERS, {"tooltip": "noise schedule"}), + "denoise": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01, "tooltip": "The amount of noise to remove. This amount is the noise added at the start, and the higher it is, the more the input latent will be modified before being returned."}), + "basic_pipe": ("BASIC_PIPE", {"tooltip": "basic_pipe input for sampling"}) + }, + "optional": { + "scheduler_func_opt": ("SCHEDULER_FUNC", {"tooltip": "[OPTIONAL] Noise schedule generation function. If this is set, the scheduler widget will be ignored."}), + } + } + + OUTPUT_TOOLTIPS = ("sampler wrapper. (Can be used when generating a regional_prompt.)",) + + RETURN_TYPES = ("KSAMPLER",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Sampler" + + @staticmethod + def doit(seed, steps, cfg, sampler_name, scheduler, denoise, basic_pipe, scheduler_func_opt=None): + model, _, _, positive, negative = basic_pipe + sampler = KSamplerWrapper(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, denoise, scheduler_func=scheduler_func_opt) + return (sampler, ) + + +class KSamplerAdvancedProvider: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0, "toolip": "classifier free guidance value"}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS, {"toolip": "sampler"}), + "scheduler": (core.SCHEDULERS, {"toolip": "noise schedule"}), + "sigma_factor": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 10.0, "step": 0.01, "toolip": "Multiplier of noise schedule"}), + "basic_pipe": ("BASIC_PIPE", {"toolip": "basic_pipe input for sampling"}) + }, + "optional": { + "sampler_opt": ("SAMPLER", {"toolip": "[OPTIONAL] Uses the passed sampler instead of internal impact_sampler."}), + "scheduler_func_opt": ("SCHEDULER_FUNC", {"toolip": "[OPTIONAL] Noise schedule generation function. If this is set, the scheduler widget will be ignored."}), + } + } + + OUTPUT_TOOLTIPS = ("sampler wrapper. (Can be used when generating a regional_prompt.)", ) + + RETURN_TYPES = ("KSAMPLER_ADVANCED",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Sampler" + + @staticmethod + def doit(cfg, sampler_name, scheduler, basic_pipe, sigma_factor=1.0, sampler_opt=None, scheduler_func_opt=None): + model, _, _, positive, negative = basic_pipe + sampler = KSamplerAdvancedWrapper(model, cfg, sampler_name, scheduler, positive, negative, sampler_opt=sampler_opt, sigma_factor=sigma_factor, scheduler_func=scheduler_func_opt) + return (sampler, ) + + +class TwoSamplersForMask: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "latent_image": ("LATENT", {"tooltip": "input latent image"}), + "base_sampler": ("KSAMPLER", {"tooltip": "Sampler to apply to the region outside the mask."}), + "mask_sampler": ("KSAMPLER", {"tooltip": "Sampler to apply to the masked region."}), + "mask": ("MASK", {"tooltip": "region mask"}) + }, + } + + OUTPUT_TOOLTIPS = ("result latent", ) + + RETURN_TYPES = ("LATENT", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Sampler" + + @staticmethod + def doit(latent_image, base_sampler, mask_sampler, mask): + inv_mask = torch.where(mask != 1.0, torch.tensor(1.0), torch.tensor(0.0)) + + latent_image['noise_mask'] = inv_mask + new_latent_image = base_sampler.sample(latent_image) + + new_latent_image['noise_mask'] = mask + new_latent_image = mask_sampler.sample(new_latent_image) + + del new_latent_image['noise_mask'] + + return (new_latent_image, ) + + +class TwoAdvancedSamplersForMask: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff, "tooltip": "Random seed to use for generating CPU noise for sampling."}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000, "tooltip": "total sampling steps"}), + "denoise": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01, "tooltip": "The amount of noise to remove. This amount is the noise added at the start, and the higher it is, the more the input latent will be modified before being returned."}), + "samples": ("LATENT", {"tooltip": "input latent image"}), + "base_sampler": ("KSAMPLER_ADVANCED", {"tooltip": "Sampler to apply to the region outside the mask."}), + "mask_sampler": ("KSAMPLER_ADVANCED", {"tooltip": "Sampler to apply to the masked region."}), + "mask": ("MASK", {"tooltip": "region mask"}), + "overlap_factor": ("INT", {"default": 10, "min": 0, "max": 10000, "tooltip": "To smooth the seams of the region boundaries, expand the mask by the overlap_factor amount to overlap with other regions."}) + }, + } + + OUTPUT_TOOLTIPS = ("result latent", ) + + RETURN_TYPES = ("LATENT", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Sampler" + + @staticmethod + def doit(seed, steps, denoise, samples, base_sampler, mask_sampler, mask, overlap_factor): + regional_prompts = RegionalPrompt().doit(mask=mask, advanced_sampler=mask_sampler)[0] + + return RegionalSampler().doit(seed=seed, seed_2nd=0, seed_2nd_mode="ignore", steps=steps, base_only_steps=1, + denoise=denoise, samples=samples, base_sampler=base_sampler, + regional_prompts=regional_prompts, overlap_factor=overlap_factor, + restore_latent=True, additional_mode="ratio between", + additional_sampler="AUTO", additional_sigma_ratio=0.3) + + +class RegionalPrompt: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "mask": ("MASK", {"tooltip": "region mask"}), + "advanced_sampler": ("KSAMPLER_ADVANCED", {"tooltip": "sampler for specified region"}), + }, + "optional": { + "variation_seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff, "tooltip": "Sets the extra seed to be used for noise variation."}), + "variation_strength": ("FLOAT", {"default": 0.0, "min": 0.0, "max": 1.0, "step": 0.01, "tooltip": "Sets the strength of the noise variation."}), + "variation_method": (["linear", "slerp"], {"tooltip": "Sets how the original noise and extra noise are blended together."}), + } + } + + OUTPUT_TOOLTIPS = ("regional prompts. (Can be used in the RegionalSampler.)", ) + + RETURN_TYPES = ("REGIONAL_PROMPTS", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Regional" + + @staticmethod + def doit(mask, advanced_sampler, variation_seed=0, variation_strength=0.0, variation_method="linear"): + regional_prompt = core.REGIONAL_PROMPT(mask, advanced_sampler, variation_seed=variation_seed, variation_strength=variation_strength, variation_method=variation_method) + return ([regional_prompt], ) + + +class CombineRegionalPrompts: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "regional_prompts1": ("REGIONAL_PROMPTS", {"tooltip": "input regional_prompts. (Connecting to the input slot increases the number of additional slots.)"}), + }, + } + + OUTPUT_TOOLTIPS = ("Combined REGIONAL_PROMPTS", ) + + RETURN_TYPES = ("REGIONAL_PROMPTS", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Regional" + + @staticmethod + def doit(**kwargs): + res = [] + for k, v in kwargs.items(): + res += v + + return (res, ) + + +class CombineConditionings: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "conditioning1": ("CONDITIONING", { "tooltip": "input conditionings. (Connecting to the input slot increases the number of additional slots.)" }), + }, + } + + OUTPUT_TOOLTIPS = ("Combined conditioning", ) + + RETURN_TYPES = ("CONDITIONING", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + @staticmethod + def doit(**kwargs): + res = [] + for k, v in kwargs.items(): + res += v + + return (res, ) + + +class ConcatConditionings: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "conditioning1": ("CONDITIONING", { "tooltip": "input conditionings. (Connecting to the input slot increases the number of additional slots.)" }), + }, + } + + OUTPUT_TOOLTIPS = ("Concatenated conditioning", ) + + RETURN_TYPES = ("CONDITIONING", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + @staticmethod + def doit(**kwargs): + conditioning_to = list(kwargs.values())[0] + + for k, conditioning_from in list(kwargs.items())[1:]: + out = [] + if len(conditioning_from) > 1: + print("Warning: ConcatConditionings {k} contains more than 1 cond, only the first one will actually be applied to conditioning1.") + + cond_from = conditioning_from[0][0] + + for i in range(len(conditioning_to)): + t1 = conditioning_to[i][0] + tw = torch.cat((t1, cond_from), 1) + n = [tw, conditioning_to[i][1].copy()] + out.append(n) + + conditioning_to = out + + return (out, ) + + +class RegionalSampler: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff, "tooltip": "Random seed to use for generating CPU noise for sampling."}), + "seed_2nd": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff, "tooltip": "Additional noise seed. The behavior is determined by seed_2nd_mode."}), + "seed_2nd_mode": (["ignore", "fixed", "seed+seed_2nd", "seed-seed_2nd", "increment", "decrement", "randomize"], {"tooltip": "application method of seed_2nd. 1) ignore: Do not use seed_2nd. In the base only sampling stage, the seed is applied as a noise seed, and in the regional sampling stage, denoising is performed as it is without additional noise. 2) Others: In the base only sampling stage, the seed is applied as a noise seed, and once it is closed so that there is no leftover noise, new noise is added with seed_2nd and the regional samping stage is performed. a) fixed: Use seed_2nd as it is as an additional noise seed. b) seed+seed_2nd: Apply the value of seed+seed_2nd as an additional noise seed. c) seed-seed_2nd: Apply the value of seed-seed_2nd as an additional noise seed. d) increment: Not implemented yet. Same with fixed. e) decrement: Not implemented yet. Same with fixed. f) randomize: Not implemented yet. Same with fixed."}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000, "tooltip": "total sampling steps"}), + "base_only_steps": ("INT", {"default": 2, "min": 0, "max": 10000, "tooltip": "total sampling steps"}), + "denoise": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01, "tooltip": "The amount of noise to remove. This amount is the noise added at the start, and the higher it is, the more the input latent will be modified before being returned."}), + "samples": ("LATENT", {"tooltip": "input latent image"}), + "base_sampler": ("KSAMPLER_ADVANCED", {"tooltip": "The sampler applied outside the area set by the regional_prompt."}), + "regional_prompts": ("REGIONAL_PROMPTS", {"tooltip": "The prompt applied to each region"}), + "overlap_factor": ("INT", {"default": 10, "min": 0, "max": 10000, "tooltip": "To smooth the seams of the region boundaries, expand the mask set in regional_prompts by the overlap_factor amount to overlap with other regions."}), + "restore_latent": ("BOOLEAN", {"default": True, "label_on": "enabled", "label_off": "disabled", "tooltip": "At each step, restore the noise outside the mask area to its original state, as per the principle of inpainting. This option is provided for backward compatibility, and it is recommended to always set it to true."}), + "additional_mode": (["DISABLE", "ratio additional", "ratio between"], {"default": "ratio between", "tooltip": "..._sde or uni_pc and other special samplers are used, the region is not properly denoised, and it causes a phenomenon that destroys the overall harmony. To compensate for this, a recovery operation is performed using another sampler. This requires a longer time for sampling because a second sampling is performed at each step in each region using a special sampler. 1) DISABLE: Disable this feature. 2) ratio additional: After performing the denoise amount to be performed in the step with the sampler set in the region, the recovery sampler is additionally applied by the additional_sigma_ratio. If you use this option, the total denoise amount increases by additional_sigma_ratio. 3) ratio between: The denoise amount to be performed in the step with the sampler set in the region and the denoise amount to be applied to the recovery sampler are divided by additional_sigma_ratio, and denoise is performed for each denoise amount. If you use this option, the total denoise amount does not change."}), + "additional_sampler": (["AUTO", "euler", "heun", "heunpp2", "dpm_2", "dpm_fast", "dpmpp_2m", "ddpm"], {"tooltip": "1) AUTO: Automatically set the recovery sampler. If the sampler is uni_pc, uni_pc_bh2, dpmpp_sde, dpmpp_sde_gpu, the dpm_fast sampler is selected If the sampler is dpmpp_2m_sde, dpmpp_2m_sde_gpu, dpmpp_3m_sde, dpmpp_3m_sde_gpu, the dpmpp_2m sampler is selected. 2) Others: Manually set the recovery sampler."}), + "additional_sigma_ratio": ("FLOAT", {"default": 0.3, "min": 0.0, "max": 1.0, "step": 0.01, "tooltip": "Multiplier of noise schedule to be applied according to additional_mode."}), + }, + "hidden": {"unique_id": "UNIQUE_ID"}, + } + + OUTPUT_TOOLTIPS = ("result latent", ) + + RETURN_TYPES = ("LATENT", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Regional" + + @staticmethod + def separated_sample(*args, **kwargs): + return separated_sample(*args, **kwargs) + + @staticmethod + def mask_erosion(samples, mask, grow_mask_by): + mask = mask.clone() + + w = samples['samples'].shape[3] + h = samples['samples'].shape[2] + + mask2 = torch.nn.functional.interpolate(mask.reshape((-1, 1, mask.shape[-2], mask.shape[-1])), size=(w, h), mode="bilinear") + if grow_mask_by == 0: + mask_erosion = mask2 + else: + kernel_tensor = torch.ones((1, 1, grow_mask_by, grow_mask_by)) + padding = math.ceil((grow_mask_by - 1) / 2) + + mask_erosion = torch.clamp(torch.nn.functional.conv2d(mask2.round(), kernel_tensor, padding=padding), 0, 1) + + return mask_erosion[:, :, :w, :h].round() + + @staticmethod + def doit(seed, seed_2nd, seed_2nd_mode, steps, base_only_steps, denoise, samples, base_sampler, regional_prompts, overlap_factor, restore_latent, + additional_mode, additional_sampler, additional_sigma_ratio, unique_id=None): + + samples = samples.copy() + samples['samples'] = comfy.sample.fix_empty_latent_channels(base_sampler.params[0], samples['samples']) + + if restore_latent: + latent_compositor = nodes.NODE_CLASS_MAPPINGS['LatentCompositeMasked']() + else: + latent_compositor = None + + masks = [regional_prompt.mask.numpy() for regional_prompt in regional_prompts] + masks = [np.ceil(mask).astype(np.int32) for mask in masks] + combined_mask = torch.from_numpy(np.bitwise_or.reduce(masks)) + + inv_mask = torch.where(combined_mask == 0, torch.tensor(1.0), torch.tensor(0.0)) + + adv_steps = int(steps / denoise) + start_at_step = adv_steps - steps + + region_len = len(regional_prompts) + total = steps*region_len + + leftover_noise = False + if base_only_steps > 0: + if seed_2nd_mode == 'ignore': + leftover_noise = True + + noise = Noise_RandomNoise(seed).generate_noise(samples) + + for rp in regional_prompts: + noise = rp.touch_noise(noise) + + samples = base_sampler.sample_advanced(True, seed, adv_steps, samples, start_at_step, start_at_step + base_only_steps, leftover_noise, recovery_mode="DISABLE", noise=noise) + + if seed_2nd_mode == "seed+seed_2nd": + seed += seed_2nd + if seed > 1125899906842624: + seed = seed - 1125899906842624 + elif seed_2nd_mode == "seed-seed_2nd": + seed -= seed_2nd + if seed < 0: + seed += 1125899906842624 + elif seed_2nd_mode != 'ignore': + seed = seed_2nd + + new_latent_image = samples.copy() + base_latent_image = None + + if not leftover_noise: + add_noise = True + noise = Noise_RandomNoise(seed).generate_noise(samples) + + for rp in regional_prompts: + noise = rp.touch_noise(noise) + else: + add_noise = False + noise = None + + for i in range(start_at_step+base_only_steps, adv_steps): + core.update_node_status(unique_id, f"{i}/{steps} steps | ", ((i-start_at_step)*region_len)/total) + + new_latent_image['noise_mask'] = inv_mask + new_latent_image = base_sampler.sample_advanced(add_noise, seed, adv_steps, new_latent_image, + start_at_step=i, end_at_step=i + 1, return_with_leftover_noise=True, + recovery_mode=additional_mode, recovery_sampler=additional_sampler, recovery_sigma_ratio=additional_sigma_ratio, noise=noise) + + if restore_latent: + if 'noise_mask' in new_latent_image: + del new_latent_image['noise_mask'] + base_latent_image = new_latent_image.copy() + + j = 1 + for regional_prompt in regional_prompts: + if restore_latent: + new_latent_image = base_latent_image.copy() + + core.update_node_status(unique_id, f"{i}/{steps} steps | {j}/{region_len}", ((i-start_at_step)*region_len + j)/total) + + region_mask = regional_prompt.get_mask_erosion(overlap_factor).squeeze(0).squeeze(0) + + new_latent_image['noise_mask'] = region_mask + new_latent_image = regional_prompt.sampler.sample_advanced(False, seed, adv_steps, new_latent_image, i, i + 1, True, + recovery_mode=additional_mode, recovery_sampler=additional_sampler, recovery_sigma_ratio=additional_sigma_ratio) + + if restore_latent: + del new_latent_image['noise_mask'] + base_latent_image = latent_compositor.composite(base_latent_image, new_latent_image, 0, 0, False, region_mask)[0] + new_latent_image = base_latent_image + + j += 1 + + add_noise = False + + # finalize + core.update_node_status(unique_id, f"finalize") + if base_latent_image is not None: + new_latent_image = base_latent_image + else: + base_latent_image = new_latent_image + + new_latent_image['noise_mask'] = inv_mask + new_latent_image = base_sampler.sample_advanced(False, seed, adv_steps, new_latent_image, adv_steps, adv_steps+1, False, + recovery_mode=additional_mode, recovery_sampler=additional_sampler, recovery_sigma_ratio=additional_sigma_ratio) + + core.update_node_status(unique_id, f"{steps}/{steps} steps", total) + core.update_node_status(unique_id, "", None) + + if restore_latent: + new_latent_image = base_latent_image + + if 'noise_mask' in new_latent_image: + del new_latent_image['noise_mask'] + + return (new_latent_image, ) + + +class RegionalSamplerAdvanced: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "add_noise": ("BOOLEAN", {"default": True, "label_on": "enabled", "label_off": "disabled", "tooltip": "Whether to add noise"}), + "noise_seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff, "tooltip": "Random seed to use for generating CPU noise for sampling."}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000, "tooltip": "total sampling steps"}), + "start_at_step": ("INT", {"default": 0, "min": 0, "max": 10000, "tooltip": "The starting step of the sampling to be applied at this node within the range of 'steps'."}), + "end_at_step": ("INT", {"default": 10000, "min": 0, "max": 10000, "tooltip": "The step at which sampling applied at this node will stop within the range of steps (if greater than steps, sampling will continue only up to steps)."}), + "overlap_factor": ("INT", {"default": 10, "min": 0, "max": 10000, "tooltip": "To smooth the seams of the region boundaries, expand the mask set in regional_prompts by the overlap_factor amount to overlap with other regions."}), + "restore_latent": ("BOOLEAN", {"default": True, "label_on": "enabled", "label_off": "disabled", "tooltip": "At each step, restore the noise outside the mask area to its original state, as per the principle of inpainting. This option is provided for backward compatibility, and it is recommended to always set it to true."}), + "return_with_leftover_noise": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled", "tooltip": "Whether to return the latent with noise remaining if the noise has not been completely removed according to the noise schedule, or to completely remove the noise before returning it."}), + "latent_image": ("LATENT", {"tooltip": "input latent image"}), + "base_sampler": ("KSAMPLER_ADVANCED", {"tooltip": "The sampler applied outside the area set by the regional_prompt."}), + "regional_prompts": ("REGIONAL_PROMPTS", {"tooltip": "The prompt applied to each region"}), + "additional_mode": (["DISABLE", "ratio additional", "ratio between"], {"default": "ratio between", "tooltip": "..._sde or uni_pc and other special samplers are used, the region is not properly denoised, and it causes a phenomenon that destroys the overall harmony. To compensate for this, a recovery operation is performed using another sampler. This requires a longer time for sampling because a second sampling is performed at each step in each region using a special sampler. 1) DISABLE: Disable this feature. 2) ratio additional: After performing the denoise amount to be performed in the step with the sampler set in the region, the recovery sampler is additionally applied by the additional_sigma_ratio. If you use this option, the total denoise amount increases by additional_sigma_ratio. 3) ratio between: The denoise amount to be performed in the step with the sampler set in the region and the denoise amount to be applied to the recovery sampler are divided by additional_sigma_ratio, and denoise is performed for each denoise amount. If you use this option, the total denoise amount does not change."}), + "additional_sampler": (["AUTO", "euler", "heun", "heunpp2", "dpm_2", "dpm_fast", "dpmpp_2m", "ddpm"], {"tooltip": "1) AUTO: Automatically set the recovery sampler. If the sampler is uni_pc, uni_pc_bh2, dpmpp_sde, dpmpp_sde_gpu, the dpm_fast sampler is selected If the sampler is dpmpp_2m_sde, dpmpp_2m_sde_gpu, dpmpp_3m_sde, dpmpp_3m_sde_gpu, the dpmpp_2m sampler is selected. 2) Others: Manually set the recovery sampler."}), + "additional_sigma_ratio": ("FLOAT", {"default": 0.3, "min": 0.0, "max": 1.0, "step": 0.01, "tooltip": "Multiplier of noise schedule to be applied according to additional_mode."}), + }, + "hidden": {"unique_id": "UNIQUE_ID"}, + } + + OUTPUT_TOOLTIPS = ("result latent", ) + + RETURN_TYPES = ("LATENT", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Regional" + + @staticmethod + def doit(add_noise, noise_seed, steps, start_at_step, end_at_step, overlap_factor, restore_latent, return_with_leftover_noise, latent_image, base_sampler, regional_prompts, + additional_mode, additional_sampler, additional_sigma_ratio, unique_id): + + new_latent_image = latent_image.copy() + new_latent_image['samples'] = comfy.sample.fix_empty_latent_channels(base_sampler.params[0], new_latent_image['samples']) + + if restore_latent: + latent_compositor = nodes.NODE_CLASS_MAPPINGS['LatentCompositeMasked']() + else: + latent_compositor = None + + masks = [regional_prompt.mask.numpy() for regional_prompt in regional_prompts] + masks = [np.ceil(mask).astype(np.int32) for mask in masks] + combined_mask = torch.from_numpy(np.bitwise_or.reduce(masks)) + + inv_mask = torch.where(combined_mask == 0, torch.tensor(1.0), torch.tensor(0.0)) + + region_len = len(regional_prompts) + end_at_step = min(steps, end_at_step) + total = (end_at_step - start_at_step) * region_len + + base_latent_image = None + region_masks = {} + + for i in range(start_at_step, end_at_step-1): + core.update_node_status(unique_id, f"{start_at_step+i}/{end_at_step} steps | ", ((i-start_at_step)*region_len)/total) + + cur_add_noise = True if i == start_at_step and add_noise else False + + if cur_add_noise: + noise = Noise_RandomNoise(noise_seed).generate_noise(new_latent_image) + for rp in regional_prompts: + noise = rp.touch_noise(noise) + else: + noise = None + + new_latent_image['noise_mask'] = inv_mask + new_latent_image = base_sampler.sample_advanced(cur_add_noise, noise_seed, steps, new_latent_image, i, i + 1, True, + recovery_mode=additional_mode, recovery_sampler=additional_sampler, recovery_sigma_ratio=additional_sigma_ratio, noise=noise) + + if restore_latent: + del new_latent_image['noise_mask'] + base_latent_image = new_latent_image.copy() + + j = 1 + for regional_prompt in regional_prompts: + if restore_latent: + new_latent_image = base_latent_image.copy() + + core.update_node_status(unique_id, f"{start_at_step+i}/{end_at_step} steps | {j}/{region_len}", ((i-start_at_step)*region_len + j)/total) + + if j not in region_masks: + region_mask = regional_prompt.get_mask_erosion(overlap_factor).squeeze(0).squeeze(0) + region_masks[j] = region_mask + else: + region_mask = region_masks[j] + + new_latent_image['noise_mask'] = region_mask + new_latent_image = regional_prompt.sampler.sample_advanced(False, noise_seed, steps, new_latent_image, i, i + 1, True, + recovery_mode=additional_mode, recovery_sampler=additional_sampler, recovery_sigma_ratio=additional_sigma_ratio) + + if restore_latent: + del new_latent_image['noise_mask'] + base_latent_image = latent_compositor.composite(base_latent_image, new_latent_image, 0, 0, False, region_mask)[0] + new_latent_image = base_latent_image + + j += 1 + + # finalize + core.update_node_status(unique_id, f"finalize") + if base_latent_image is not None: + new_latent_image = base_latent_image + else: + base_latent_image = new_latent_image + + new_latent_image['noise_mask'] = inv_mask + new_latent_image = base_sampler.sample_advanced(False, noise_seed, steps, new_latent_image, end_at_step-1, end_at_step, return_with_leftover_noise, + recovery_mode=additional_mode, recovery_sampler=additional_sampler, recovery_sigma_ratio=additional_sigma_ratio) + + core.update_node_status(unique_id, f"{end_at_step}/{end_at_step} steps", total) + core.update_node_status(unique_id, "", None) + + if restore_latent: + new_latent_image = base_latent_image + + if 'noise_mask' in new_latent_image: + del new_latent_image['noise_mask'] + + return (new_latent_image, ) + + +class KSamplerBasicPipe: + @classmethod + def INPUT_TYPES(s): + return {"required": + {"basic_pipe": ("BASIC_PIPE", {"tooltip": "basic_pipe input for sampling"}), + "seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff, "tooltip": "Random seed to use for generating CPU noise for sampling."}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000, "tooltip": "total sampling steps"}), + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0, "tooltip": "classifier free guidance value"}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS, {"tooltip": "sampler"}), + "scheduler": (core.SCHEDULERS, {"tooltip": "noise schedule"}), + "latent_image": ("LATENT", {"tooltip": "input latent image"}), + "denoise": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01, "tooltip": "The amount of noise to remove. This amount is the noise added at the start, and the higher it is, the more the input latent will be modified before being returned."}), + }, + "optional": + { + "scheduler_func_opt": ("SCHEDULER_FUNC", {"tooltip": "[OPTIONAL] Noise schedule generation function. If this is set, the scheduler widget will be ignored."}), + } + } + + OUTPUT_TOOLTIPS = ("passthrough input basic_pipe", "result latent", "VAE in basic_pipe") + + RETURN_TYPES = ("BASIC_PIPE", "LATENT", "VAE") + FUNCTION = "sample" + + CATEGORY = "ImpactPack/sampling" + + @staticmethod + def sample(basic_pipe, seed, steps, cfg, sampler_name, scheduler, latent_image, denoise=1.0, scheduler_func_opt=None): + model, clip, vae, positive, negative = basic_pipe + latent = impact_sample(model, seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, denoise, scheduler_func=scheduler_func_opt) + return basic_pipe, latent, vae + + +class KSamplerAdvancedBasicPipe: + @classmethod + def INPUT_TYPES(s): + return {"required": + {"basic_pipe": ("BASIC_PIPE", {"tooltip": "basic_pipe input for sampling"}), + "add_noise": ("BOOLEAN", {"default": True, "label_on": "enable", "label_off": "disable", "tooltip": "Whether to add noise"}), + "noise_seed": ("INT", {"default": 0, "min": 0, "max": 0xffffffffffffffff, "tooltip": "Random seed to use for generating CPU noise for sampling."}), + "steps": ("INT", {"default": 20, "min": 1, "max": 10000, "tooltip": "total sampling steps"}), + "cfg": ("FLOAT", {"default": 8.0, "min": 0.0, "max": 100.0, "tooltip": "classifier free guidance value"}), + "sampler_name": (comfy.samplers.KSampler.SAMPLERS, {"tooltip": "sampler"}), + "scheduler": (core.SCHEDULERS, {"tooltip": "noise schedule"}), + "latent_image": ("LATENT", {"tooltip": "input latent image"}), + "start_at_step": ("INT", {"default": 0, "min": 0, "max": 10000, "tooltip": "The starting step of the sampling to be applied at this node within the range of 'steps'."}), + "end_at_step": ("INT", {"default": 10000, "min": 0, "max": 10000, "tooltip": "The step at which sampling applied at this node will stop within the range of steps (if greater than steps, sampling will continue only up to steps)."}), + "return_with_leftover_noise": ("BOOLEAN", {"default": False, "label_on": "enable", "label_off": "disable", "tooltip": "Whether to return the latent with noise remaining if the noise has not been completely removed according to the noise schedule, or to completely remove the noise before returning it."}), + }, + "optional": + { + "scheduler_func_opt": ("SCHEDULER_FUNC", {"tooltip": "[OPTIONAL] Noise schedule generation function. If this is set, the scheduler widget will be ignored."}), + } + } + + OUTPUT_TOOLTIPS = ("passthrough input basic_pipe", "result latent", "VAE in basic_pipe") + + RETURN_TYPES = ("BASIC_PIPE", "LATENT", "VAE") + FUNCTION = "sample" + + CATEGORY = "ImpactPack/sampling" + + @staticmethod + def sample(basic_pipe, add_noise, noise_seed, steps, cfg, sampler_name, scheduler, latent_image, start_at_step, end_at_step, return_with_leftover_noise, denoise=1.0, scheduler_func_opt=None): + model, clip, vae, positive, negative = basic_pipe + + latent = separated_sample(model, add_noise, noise_seed, steps, cfg, sampler_name, scheduler, positive, negative, latent_image, start_at_step, end_at_step, return_with_leftover_noise, scheduler_func=scheduler_func_opt) + return basic_pipe, latent, vae + + +class GITSSchedulerFuncProvider: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "coeff": ("FLOAT", {"default": 1.20, "min": 0.80, "max": 1.50, "step": 0.05, "tooltip": "coeff factor of GITS Scheduler"}), + "denoise": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.01, "tooltip": "denoise amount for noise schedule"}), + } + } + + OUTPUT_TOOLTIPS = ("Returns a function that generates a noise schedule using GITSScheduler. This can be used in place of a predetermined noise schedule to dynamically generate a noise schedule based on the steps.",) + + RETURN_TYPES = ("SCHEDULER_FUNC",) + CATEGORY = "ImpactPack/sampling" + + FUNCTION = "doit" + + @staticmethod + def doit(coeff, denoise): + def f(model, sampler, steps): + if 'GITSScheduler' not in nodes.NODE_CLASS_MAPPINGS: + raise Exception("[Impact Pack] ComfyUI is an outdated version. Cannot use GITSScheduler.") + + scheduler = nodes.NODE_CLASS_MAPPINGS['GITSScheduler']() + return scheduler.get_sigmas(coeff, steps, denoise)[0] + + return (f, ) + + +class NegativeConditioningPlaceholder: + @classmethod + def INPUT_TYPES(s): + return {"required": {}} + + OUTPUT_TOOLTIPS = ("This is a Placeholder for the FLUX model that does not use Negative Conditioning.",) + + RETURN_TYPES = ("CONDITIONING",) + CATEGORY = "ImpactPack/sampling" + + FUNCTION = "doit" + + @staticmethod + def doit(): + return ("NegativePlaceholder", ) diff --git a/ComfyUI-Impact-Pack/modules/impact/util_nodes.py b/ComfyUI-Impact-Pack/modules/impact/util_nodes.py new file mode 100644 index 0000000000000000000000000000000000000000..3bad6325610cfce63ab4a5cc45d9c3d1f6f8c13b --- /dev/null +++ b/ComfyUI-Impact-Pack/modules/impact/util_nodes.py @@ -0,0 +1,702 @@ +from impact.utils import any_typ, ByPassTypeTuple, make_3d_mask +import comfy_extras.nodes_mask +from nodes import MAX_RESOLUTION +import torch +import comfy +import sys +import nodes +import re +import impact.core as core +from server import PromptServer +import inspect + + +class GeneralSwitch: + @classmethod + def INPUT_TYPES(s): + dyn_inputs = {"input1": (any_typ, {"lazy": True, "tooltip": "Any input. When connected, one more input slot is added."}), } + if core.is_execution_model_version_supported(): + stack = inspect.stack() + if stack[2].function == 'get_input_info': + # bypass validation + class AllContainer: + def __contains__(self, item): + return True + + def __getitem__(self, key): + return any_typ, {"lazy": True} + + dyn_inputs = AllContainer() + + inputs = {"required": { + "select": ("INT", {"default": 1, "min": 1, "max": 999999, "step": 1, "tooltip": "The input number you want to output among the inputs"}), + "sel_mode": ("BOOLEAN", {"default": False, "label_on": "select_on_prompt", "label_off": "select_on_execution", "forceInput": False, + "tooltip": "In the case of 'select_on_execution', the selection is dynamically determined at the time of workflow execution. 'select_on_prompt' is an option that exists for older versions of ComfyUI, and it makes the decision before the workflow execution."}), + }, + "optional": dyn_inputs, + "hidden": {"unique_id": "UNIQUE_ID", "extra_pnginfo": "EXTRA_PNGINFO"} + } + + return inputs + + RETURN_TYPES = (any_typ, "STRING", "INT") + RETURN_NAMES = ("selected_value", "selected_label", "selected_index") + OUTPUT_TOOLTIPS = ("Output is generated only from the input chosen by the 'select' value.", "Slot label of the selected input slot", "Outputs the select value as is") + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def check_lazy_status(self, *args, **kwargs): + selected_index = int(kwargs['select']) + input_name = f"input{selected_index}" + + print(f"SELECTED: {input_name}") + + return [input_name] + + @staticmethod + def doit(*args, **kwargs): + selected_index = int(kwargs['select']) + input_name = f"input{selected_index}" + + selected_label = input_name + node_id = kwargs['unique_id'] + + if 'extra_pnginfo' in kwargs and kwargs['extra_pnginfo'] is not None: + nodelist = kwargs['extra_pnginfo']['workflow']['nodes'] + for node in nodelist: + if str(node['id']) == node_id: + inputs = node['inputs'] + + for slot in inputs: + if slot['name'] == input_name and 'label' in slot: + selected_label = slot['label'] + + break + else: + print(f"[Impact-Pack] The switch node does not guarantee proper functioning in API mode.") + + if input_name in kwargs: + return kwargs[input_name], selected_label, selected_index + else: + print(f"ImpactSwitch: invalid select index (ignored)") + return None, "", selected_index + +class LatentSwitch: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "select": ("INT", {"default": 1, "min": 1, "max": 99999, "step": 1}), + "latent1": ("LATENT",), + }, + } + + RETURN_TYPES = ("LATENT", ) + + OUTPUT_NODE = True + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, *args, **kwargs): + input_name = f"latent{int(kwargs['select'])}" + + if input_name in kwargs: + return (kwargs[input_name],) + else: + print(f"LatentSwitch: invalid select index ('latent1' is selected)") + return (kwargs['latent1'],) + + +class ImageMaskSwitch: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "select": ("INT", {"default": 1, "min": 1, "max": 4, "step": 1}), + "images1": ("IMAGE",), + }, + + "optional": { + "mask1_opt": ("MASK",), + "images2_opt": ("IMAGE",), + "mask2_opt": ("MASK",), + "images3_opt": ("IMAGE",), + "mask3_opt": ("MASK",), + "images4_opt": ("IMAGE",), + "mask4_opt": ("MASK",), + }, + } + + RETURN_TYPES = ("IMAGE", "MASK",) + + OUTPUT_NODE = True + + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, select, images1, mask1_opt=None, images2_opt=None, mask2_opt=None, images3_opt=None, mask3_opt=None, + images4_opt=None, mask4_opt=None): + if select == 1: + return images1, mask1_opt, + elif select == 2: + return images2_opt, mask2_opt, + elif select == 3: + return images3_opt, mask3_opt, + else: + return images4_opt, mask4_opt, + + +class GeneralInversedSwitch: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "select": ("INT", {"default": 1, "min": 1, "max": 999999, "step": 1, "tooltip": "The output number you want to send from the input"}), + "input": (any_typ, {"tooltip": "Any input. When connected, one more input slot is added."}), + + }, + "optional": { + "sel_mode": ("BOOLEAN", {"default": False, "label_on": "select_on_prompt", "label_off": "select_on_execution", "forceInput": False, + "tooltip": "In the case of 'select_on_execution', the selection is dynamically determined at the time of workflow execution. 'select_on_prompt' is an option that exists for older versions of ComfyUI, and it makes the decision before the workflow execution."}), + }, + "hidden": {"prompt": "PROMPT", "unique_id": "UNIQUE_ID"}, + } + + RETURN_TYPES = ByPassTypeTuple((any_typ, )) + OUTPUT_TOOLTIPS = ("Output occurs only from the output selected by the 'select' value.\nWhen slots are connected, additional slots are created.", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, select, prompt, unique_id, input, **kwargs): + if core.is_execution_model_version_supported(): + from comfy_execution.graph import ExecutionBlocker + else: + print("[Impact Pack] InversedSwitch: ComfyUI is outdated. The 'select_on_execution' mode cannot function properly.") + + res = [] + + # search max output count in prompt + cnt = 0 + for x in prompt.values(): + for y in x.get('inputs', {}).values(): + if isinstance(y, list) and len(y) == 2: + if y[0] == unique_id: + cnt = max(cnt, y[1]) + + for i in range(0, cnt + 1): + if select == i+1: + res.append(input) + elif core.is_execution_model_version_supported(): + res.append(ExecutionBlocker(None)) + else: + res.append(None) + + return res + + +class RemoveNoiseMask: + @classmethod + def INPUT_TYPES(s): + return {"required": {"samples": ("LATENT",)}} + + RETURN_TYPES = ("LATENT",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, samples): + res = {key: value for key, value in samples.items() if key != 'noise_mask'} + return (res, ) + + +class ImagePasteMasked: + @classmethod + def INPUT_TYPES(s): + return { + "required": { + "destination": ("IMAGE",), + "source": ("IMAGE",), + "x": ("INT", {"default": 0, "min": 0, "max": MAX_RESOLUTION, "step": 1}), + "y": ("INT", {"default": 0, "min": 0, "max": MAX_RESOLUTION, "step": 1}), + "resize_source": ("BOOLEAN", {"default": False}), + }, + "optional": { + "mask": ("MASK",), + } + } + RETURN_TYPES = ("IMAGE",) + FUNCTION = "composite" + + CATEGORY = "image" + + def composite(self, destination, source, x, y, resize_source, mask = None): + destination = destination.clone().movedim(-1, 1) + output = comfy_extras.nodes_mask.composite(destination, source.movedim(-1, 1), x, y, mask, 1, resize_source).movedim(1, -1) + return (output,) + + +from impact.utils import any_typ + +class ImpactLogger: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "data": (any_typ,), + "text": ("STRING", {"multiline": True}), + }, + "hidden": {"prompt": "PROMPT", "extra_pnginfo": "EXTRA_PNGINFO", "unique_id": "UNIQUE_ID"}, + } + + CATEGORY = "ImpactPack/Debug" + + OUTPUT_NODE = True + + RETURN_TYPES = () + FUNCTION = "doit" + + def doit(self, data, text, prompt, extra_pnginfo, unique_id): + shape = "" + if hasattr(data, "shape"): + shape = f"{data.shape} / " + + print(f"[IMPACT LOGGER]: {shape}{data}") + + print(f" PROMPT: {prompt}") + + # for x in prompt: + # if 'inputs' in x and 'populated_text' in x['inputs']: + # print(f"PROMPT: {x['10']['inputs']['populated_text']}") + # + # for x in extra_pnginfo['workflow']['nodes']: + # if x['type'] == 'ImpactWildcardProcessor': + # print(f" WV : {x['widgets_values'][1]}\n") + + PromptServer.instance.send_sync("impact-node-feedback", {"node_id": unique_id, "widget_name": "text", "type": "TEXT", "value": f"{data}"}) + return {} + + +class ImpactDummyInput: + @classmethod + def INPUT_TYPES(s): + return {"required": {}} + + CATEGORY = "ImpactPack/Debug" + + RETURN_TYPES = (any_typ,) + FUNCTION = "doit" + + def doit(self): + return ("DUMMY",) + + +class MasksToMaskList: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "masks": ("MASK", ), + } + } + + RETURN_TYPES = ("MASK", ) + OUTPUT_IS_LIST = (True, ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Operation" + + def doit(self, masks): + if masks is None: + empty_mask = torch.zeros((64, 64), dtype=torch.float32, device="cpu") + return ([empty_mask], ) + + res = [] + + for mask in masks: + res.append(mask) + + print(f"mask len: {len(res)}") + + res = [make_3d_mask(x) for x in res] + + return (res, ) + + +class MaskListToMaskBatch: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "mask": ("MASK", ), + } + } + + INPUT_IS_LIST = True + + RETURN_TYPES = ("MASK", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Operation" + + def doit(self, mask): + if len(mask) == 1: + mask = make_3d_mask(mask[0]) + return (mask,) + elif len(mask) > 1: + mask1 = make_3d_mask(mask[0]) + + for mask2 in mask[1:]: + mask2 = make_3d_mask(mask2) + if mask1.shape[1:] != mask2.shape[1:]: + mask2 = comfy.utils.common_upscale(mask2.movedim(-1, 1), mask1.shape[2], mask1.shape[1], "lanczos", "center").movedim(1, -1) + mask1 = torch.cat((mask1, mask2), dim=0) + + return (mask1,) + else: + empty_mask = torch.zeros((1, 64, 64), dtype=torch.float32, device="cpu").unsqueeze(0) + return (empty_mask,) + + +class ImageListToImageBatch: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "images": ("IMAGE", ), + } + } + + INPUT_IS_LIST = True + + RETURN_TYPES = ("IMAGE", ) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Operation" + + def doit(self, images): + if len(images) <= 1: + return (images[0],) + else: + image1 = images[0] + for image2 in images[1:]: + if image1.shape[1:] != image2.shape[1:]: + image2 = comfy.utils.common_upscale(image2.movedim(-1, 1), image1.shape[2], image1.shape[1], "lanczos", "center").movedim(1, -1) + image1 = torch.cat((image1, image2), dim=0) + return (image1,) + + +class ImageBatchToImageList: + @classmethod + def INPUT_TYPES(s): + return {"required": {"image": ("IMAGE",), }} + + RETURN_TYPES = ("IMAGE",) + OUTPUT_IS_LIST = (True,) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, image): + images = [image[i:i + 1, ...] for i in range(image.shape[0])] + return (images, ) + + +class MakeAnyList: + @classmethod + def INPUT_TYPES(s): + return { + "required": {}, + "optional": {"value1": (any_typ,), } + } + + RETURN_TYPES = (any_typ,) + OUTPUT_IS_LIST = (True,) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, **kwargs): + values = [] + + for k, v in kwargs.items(): + if v is not None: + values.append(v) + + return (values, ) + + +class MakeMaskList: + @classmethod + def INPUT_TYPES(s): + return {"required": {"mask1": ("MASK",), }} + + RETURN_TYPES = ("MASK",) + OUTPUT_IS_LIST = (True,) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, **kwargs): + masks = [] + + for k, v in kwargs.items(): + masks.append(v) + + return (masks, ) + + +class MakeImageList: + @classmethod + def INPUT_TYPES(s): + return {"required": {"image1": ("IMAGE",), }} + + RETURN_TYPES = ("IMAGE",) + OUTPUT_IS_LIST = (True,) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, **kwargs): + images = [] + + for k, v in kwargs.items(): + images.append(v) + + return (images, ) + + +class MakeImageBatch: + @classmethod + def INPUT_TYPES(s): + return {"required": {"image1": ("IMAGE",), }} + + RETURN_TYPES = ("IMAGE",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, **kwargs): + image1 = kwargs['image1'] + del kwargs['image1'] + images = [value for value in kwargs.values()] + + if len(images) == 0: + return (image1,) + else: + for image2 in images: + if image1.shape[1:] != image2.shape[1:]: + image2 = comfy.utils.common_upscale(image2.movedim(-1, 1), image1.shape[2], image1.shape[1], "lanczos", "center").movedim(1, -1) + image1 = torch.cat((image1, image2), dim=0) + return (image1,) + + +class MakeMaskBatch: + @classmethod + def INPUT_TYPES(s): + return {"required": {"mask1": ("MASK",), }} + + RETURN_TYPES = ("MASK",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, **kwargs): + mask1 = kwargs['mask1'] + del kwargs['mask1'] + masks = [utils.make_3d_mask(value) for value in kwargs.values()] + + if len(masks) == 0: + return (mask1,) + else: + for mask2 in masks: + if mask1.shape[1:] != mask2.shape[1:]: + mask2 = comfy.utils.common_upscale(mask2.movedim(-1, 1), mask1.shape[2], mask1.shape[1], "lanczos", "center").movedim(1, -1) + mask1 = torch.cat((mask1, mask2), dim=0) + return (mask1,) + + +class ReencodeLatent: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "samples": ("LATENT", ), + "tile_mode": (["None", "Both", "Decode(input) only", "Encode(output) only"],), + "input_vae": ("VAE", ), + "output_vae": ("VAE", ), + "tile_size": ("INT", {"default": 512, "min": 320, "max": 4096, "step": 64}), + }, + } + + CATEGORY = "ImpactPack/Util" + + RETURN_TYPES = ("LATENT", ) + FUNCTION = "doit" + + def doit(self, samples, tile_mode, input_vae, output_vae, tile_size=512): + if tile_mode in ["Both", "Decode(input) only"]: + pixels = nodes.VAEDecodeTiled().decode(input_vae, samples, tile_size)[0] + else: + pixels = nodes.VAEDecode().decode(input_vae, samples)[0] + + if tile_mode in ["Both", "Encode(output) only"]: + return nodes.VAEEncodeTiled().encode(output_vae, pixels, tile_size) + else: + return nodes.VAEEncode().encode(output_vae, pixels) + + +class ReencodeLatentPipe: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "samples": ("LATENT", ), + "tile_mode": (["None", "Both", "Decode(input) only", "Encode(output) only"],), + "input_basic_pipe": ("BASIC_PIPE", ), + "output_basic_pipe": ("BASIC_PIPE", ), + }, + } + + CATEGORY = "ImpactPack/Util" + + RETURN_TYPES = ("LATENT", ) + FUNCTION = "doit" + + def doit(self, samples, tile_mode, input_basic_pipe, output_basic_pipe): + _, _, input_vae, _, _ = input_basic_pipe + _, _, output_vae, _, _ = output_basic_pipe + return ReencodeLatent().doit(samples, tile_mode, input_vae, output_vae) + + +class StringSelector: + @classmethod + def INPUT_TYPES(s): + return {"required": { + "strings": ("STRING", {"multiline": True}), + "multiline": ("BOOLEAN", {"default": False, "label_on": "enabled", "label_off": "disabled"}), + "select": ("INT", {"min": 0, "max": sys.maxsize, "step": 1, "default": 0}), + }} + + RETURN_TYPES = ("STRING",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, strings, multiline, select): + lines = strings.split('\n') + + if multiline: + result = [] + current_string = "" + + for line in lines: + if line.startswith("#"): + if current_string: + result.append(current_string.strip()) + current_string = "" + current_string += line + "\n" + + if current_string: + result.append(current_string.strip()) + + if len(result) == 0: + selected = strings + else: + selected = result[select % len(result)] + + if selected.startswith('#'): + selected = selected[1:] + else: + if len(lines) == 0: + selected = strings + else: + selected = lines[select % len(lines)] + + return (selected, ) + + +class StringListToString: + @classmethod + def INPUT_TYPES(s): + return { + "required": { + "join_with": ("STRING", {"default": "\\n"}), + "string_list": ("STRING", {"forceInput": True}), + } + } + + INPUT_IS_LIST = True + RETURN_TYPES = ("STRING",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, join_with, string_list): + # convert \\n to newline character + if join_with[0] == "\\n": + join_with[0] = "\n" + + joined_text = join_with[0].join(string_list) + + return (joined_text,) + + +class WildcardPromptFromString: + @classmethod + def INPUT_TYPES(s): + return { + "required": { + "string": ("STRING", {"forceInput": True}), + "delimiter": ("STRING", {"multiline": False, "default": "\\n" }), + "prefix_all": ("STRING", {"multiline": False}), + "postfix_all": ("STRING", {"multiline": False}), + "restrict_to_tags": ("STRING", {"multiline": False}), + "exclude_tags": ("STRING", {"multiline": False}) + }, + } + + RETURN_TYPES = ("STRING", "STRING",) + RETURN_NAMES = ("wildcard", "segs_labels",) + FUNCTION = "doit" + + CATEGORY = "ImpactPack/Util" + + def doit(self, string, delimiter, prefix_all, postfix_all, restrict_to_tags, exclude_tags): + # convert \\n to newline character + if delimiter == "\\n": + delimiter = "\n" + + # some sanity checks and normalization for later processing + if prefix_all is None: + prefix_all = "" + if postfix_all is None: + postfix_all = "" + if restrict_to_tags is None: + restrict_to_tags = "" + if exclude_tags is None: + exclude_tags = "" + + restrict_to_tags = restrict_to_tags.split(", ") + exclude_tags = exclude_tags.split(", ") + + # build the wildcard prompt per list entry + output = ["[LAB]"] + labels = [] + for x in string.split(delimiter): + label = str(len(labels) + 1) + labels.append(label) + x = x.split(", ") + # restrict to tags + if restrict_to_tags != [""]: + x = list(set(x) & set(restrict_to_tags)) + # remove tags + if exclude_tags != [""]: + x = list(set(x) - set(exclude_tags)) + # next row: