import torch import torch.nn as nn import torch.nn.functional as F from layer import conv1x1, conv3x3, BasicBlock class ResidualBlock(nn.Module): def __init__(self, in_planes, planes, norm_fn='group', stride=1): super(ResidualBlock, self).__init__() self.conv1 = nn.Conv2d(in_planes, planes, kernel_size=3, padding=1, stride=stride) self.conv2 = nn.Conv2d(planes, planes, kernel_size=3, padding=1) self.relu = nn.ReLU(inplace=True) num_groups = planes // 8 if norm_fn == 'group': self.norm1 = nn.GroupNorm(num_groups=num_groups, num_channels=planes) self.norm2 = nn.GroupNorm(num_groups=num_groups, num_channels=planes) if not stride == 1: self.norm3 = nn.GroupNorm(num_groups=num_groups, num_channels=planes) elif norm_fn == 'batch': self.norm1 = nn.BatchNorm2d(planes) self.norm2 = nn.BatchNorm2d(planes) if not stride == 1: self.norm3 = nn.BatchNorm2d(planes) elif norm_fn == 'instance': self.norm1 = nn.InstanceNorm2d(planes) self.norm2 = nn.InstanceNorm2d(planes) if not stride == 1: self.norm3 = nn.InstanceNorm2d(planes) elif norm_fn == 'none': self.norm1 = nn.Sequential() self.norm2 = nn.Sequential() if not stride == 1: self.norm3 = nn.Sequential() if stride == 1: self.downsample = None else: self.downsample = nn.Sequential( nn.Conv2d(in_planes, planes, kernel_size=1, stride=stride), self.norm3) def forward(self, x): y = x y = self.relu(self.norm1(self.conv1(y))) y = self.relu(self.norm2(self.conv2(y))) if self.downsample is not None: x = self.downsample(x) return self.relu(x+y) class BottleneckBlock(nn.Module): def __init__(self, in_planes, planes, norm_fn='group', stride=1): super(BottleneckBlock, self).__init__() self.conv1 = nn.Conv2d(in_planes, planes//4, kernel_size=1, padding=0) self.conv2 = nn.Conv2d(planes//4, planes//4, kernel_size=3, padding=1, stride=stride) self.conv3 = nn.Conv2d(planes//4, planes, kernel_size=1, padding=0) self.relu = nn.ReLU(inplace=True) num_groups = planes // 8 if norm_fn == 'group': self.norm1 = nn.GroupNorm(num_groups=num_groups, num_channels=planes//4) self.norm2 = nn.GroupNorm(num_groups=num_groups, num_channels=planes//4) self.norm3 = nn.GroupNorm(num_groups=num_groups, num_channels=planes) if not stride == 1: self.norm4 = nn.GroupNorm(num_groups=num_groups, num_channels=planes) elif norm_fn == 'batch': self.norm1 = nn.BatchNorm2d(planes//4) self.norm2 = nn.BatchNorm2d(planes//4) self.norm3 = nn.BatchNorm2d(planes) if not stride == 1: self.norm4 = nn.BatchNorm2d(planes) elif norm_fn == 'instance': self.norm1 = nn.InstanceNorm2d(planes//4) self.norm2 = nn.InstanceNorm2d(planes//4) self.norm3 = nn.InstanceNorm2d(planes) if not stride == 1: self.norm4 = nn.InstanceNorm2d(planes) elif norm_fn == 'none': self.norm1 = nn.Sequential() self.norm2 = nn.Sequential() self.norm3 = nn.Sequential() if not stride == 1: self.norm4 = nn.Sequential() if stride == 1: self.downsample = None else: self.downsample = nn.Sequential( nn.Conv2d(in_planes, planes, kernel_size=1, stride=stride), self.norm4) def forward(self, x): y = x y = self.relu(self.norm1(self.conv1(y))) y = self.relu(self.norm2(self.conv2(y))) y = self.relu(self.norm3(self.conv3(y))) if self.downsample is not None: x = self.downsample(x) return self.relu(x+y) class BasicEncoder(nn.Module): def __init__(self, output_dim=128, norm_fn='batch', dropout=0.0): super(BasicEncoder, self).__init__() self.norm_fn = norm_fn if self.norm_fn == 'group': self.norm1 = nn.GroupNorm(num_groups=8, num_channels=64) elif self.norm_fn == 'batch': self.norm1 = nn.BatchNorm2d(64) elif self.norm_fn == 'instance': self.norm1 = nn.InstanceNorm2d(64) elif self.norm_fn == 'none': self.norm1 = nn.Sequential() self.conv1 = nn.Conv2d(3, 64, kernel_size=7, stride=2, padding=3) self.relu1 = nn.ReLU(inplace=True) self.in_planes = 64 self.layer1 = self._make_layer(64, stride=1) self.layer2 = self._make_layer(96, stride=2) self.layer3 = self._make_layer(128, stride=2) # output convolution self.conv2 = nn.Conv2d(128, output_dim, kernel_size=1) self.dropout = None if dropout > 0: self.dropout = nn.Dropout2d(p=dropout) for m in self.modules(): if isinstance(m, nn.Conv2d): nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu') elif isinstance(m, (nn.BatchNorm2d, nn.InstanceNorm2d, nn.GroupNorm)): if m.weight is not None: nn.init.constant_(m.weight, 1) if m.bias is not None: nn.init.constant_(m.bias, 0) def _make_layer(self, dim, stride=1): layer1 = ResidualBlock(self.in_planes, dim, self.norm_fn, stride=stride) layer2 = ResidualBlock(dim, dim, self.norm_fn, stride=1) layers = (layer1, layer2) self.in_planes = dim return nn.Sequential(*layers) def forward(self, x): # if input is list, combine batch dimension is_list = isinstance(x, tuple) or isinstance(x, list) if is_list: batch_dim = x[0].shape[0] x = torch.cat(x, dim=0) x = self.conv1(x) x = self.norm1(x) x = self.relu1(x) x = self.layer1(x) x = self.layer2(x) x = self.layer3(x) x = self.conv2(x) if self.training and self.dropout is not None: x = self.dropout(x) if is_list: x = torch.split(x, [batch_dim, batch_dim], dim=0) return x class SmallEncoder(nn.Module): def __init__(self, output_dim=128, norm_fn='batch', dropout=0.0): super(SmallEncoder, self).__init__() self.norm_fn = norm_fn if self.norm_fn == 'group': self.norm1 = nn.GroupNorm(num_groups=8, num_channels=32) elif self.norm_fn == 'batch': self.norm1 = nn.BatchNorm2d(32) elif self.norm_fn == 'instance': self.norm1 = nn.InstanceNorm2d(32) elif self.norm_fn == 'none': self.norm1 = nn.Sequential() self.conv1 = nn.Conv2d(3, 32, kernel_size=7, stride=2, padding=3) self.relu1 = nn.ReLU(inplace=True) self.in_planes = 32 self.layer1 = self._make_layer(32, stride=1) self.layer2 = self._make_layer(64, stride=2) self.layer3 = self._make_layer(96, stride=2) self.dropout = None if dropout > 0: self.dropout = nn.Dropout2d(p=dropout) self.conv2 = nn.Conv2d(96, output_dim, kernel_size=1) for m in self.modules(): if isinstance(m, nn.Conv2d): nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu') elif isinstance(m, (nn.BatchNorm2d, nn.InstanceNorm2d, nn.GroupNorm)): if m.weight is not None: nn.init.constant_(m.weight, 1) if m.bias is not None: nn.init.constant_(m.bias, 0) def _make_layer(self, dim, stride=1): layer1 = BottleneckBlock(self.in_planes, dim, self.norm_fn, stride=stride) layer2 = BottleneckBlock(dim, dim, self.norm_fn, stride=1) layers = (layer1, layer2) self.in_planes = dim return nn.Sequential(*layers) def forward(self, x): # if input is list, combine batch dimension is_list = isinstance(x, tuple) or isinstance(x, list) if is_list: batch_dim = x[0].shape[0] x = torch.cat(x, dim=0) x = self.conv1(x) x = self.norm1(x) x = self.relu1(x) x = self.layer1(x) x = self.layer2(x) x = self.layer3(x) x = self.conv2(x) if self.training and self.dropout is not None: x = self.dropout(x) if is_list: x = torch.split(x, [batch_dim, batch_dim], dim=0) return x class ResNetFPN(nn.Module): """ ResNet18, output resolution is 1/8. Each block has 2 layers. """ def __init__(self, args, input_dim=3, output_dim=256, ratio=1.0, norm_layer=nn.BatchNorm2d, init_weight=False): super().__init__() # Config block = BasicBlock block_dims = args.block_dims initial_dim = args.initial_dim self.init_weight = init_weight self.input_dim = input_dim # Class Variable self.in_planes = initial_dim for i in range(len(block_dims)): block_dims[i] = int(block_dims[i] * ratio) # Networks self.conv1 = nn.Conv2d(input_dim, initial_dim, kernel_size=7, stride=2, padding=3) self.bn1 = norm_layer(initial_dim) self.relu = nn.ReLU(inplace=True) if args.pretrain == 'resnet34': n_block = [3, 4, 6] elif args.pretrain == 'resnet18': n_block = [2, 2, 2] else: raise NotImplementedError self.layer1 = self._make_layer(block, block_dims[0], stride=1, norm_layer=norm_layer, num=n_block[0]) # 1/2 self.layer2 = self._make_layer(block, block_dims[1], stride=2, norm_layer=norm_layer, num=n_block[1]) # 1/4 self.layer3 = self._make_layer(block, block_dims[2], stride=2, norm_layer=norm_layer, num=n_block[2]) # 1/8 self.final_conv = conv1x1(block_dims[2], output_dim) self._init_weights(args) def _init_weights(self, args): for m in self.modules(): if isinstance(m, nn.Conv2d): nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu') if m.bias is not None: nn.init.constant_(m.bias, 0) elif isinstance(m, (nn.BatchNorm2d, nn.InstanceNorm2d, nn.GroupNorm)): if m.weight is not None: nn.init.constant_(m.weight, 1) if m.bias is not None: nn.init.constant_(m.bias, 0) if self.init_weight: from torchvision.models import resnet18, ResNet18_Weights, resnet34, ResNet34_Weights if args.pretrain == 'resnet18': pretrained_dict = resnet18(weights=ResNet18_Weights.IMAGENET1K_V1).state_dict() else: pretrained_dict = resnet34(weights=ResNet34_Weights.IMAGENET1K_V1).state_dict() model_dict = self.state_dict() pretrained_dict = {k: v for k, v in pretrained_dict.items() if k in model_dict} if self.input_dim == 6: for k, v in pretrained_dict.items(): if k == 'conv1.weight': pretrained_dict[k] = torch.cat((v, v), dim=1) model_dict.update(pretrained_dict) self.load_state_dict(model_dict, strict=False) def _make_layer(self, block, dim, stride=1, norm_layer=nn.BatchNorm2d, num=2): layers = [] layers.append(block(self.in_planes, dim, stride=stride, norm_layer=norm_layer)) for i in range(num - 1): layers.append(block(dim, dim, stride=1, norm_layer=norm_layer)) self.in_planes = dim return nn.Sequential(*layers) def forward(self, x): # ResNet Backbone x = self.relu(self.bn1(self.conv1(x))) for i in range(len(self.layer1)): x = self.layer1[i](x) for i in range(len(self.layer2)): x = self.layer2[i](x) for i in range(len(self.layer3)): x = self.layer3[i](x) # Output output = self.final_conv(x) return output