Spaces:
Runtime error
Runtime error
jhj0517
commited on
Commit
·
3423be2
1
Parent(s):
62a455e
Add create_filtered_video() and move logic to the inference class
Browse files- modules/sam_inference.py +26 -15
modules/sam_inference.py
CHANGED
@@ -23,7 +23,8 @@ from modules.mask_utils import (
|
|
23 |
create_mask_pixelized_image,
|
24 |
create_solid_color_mask_image
|
25 |
)
|
26 |
-
from modules.video_utils import get_frames_from_dir
|
|
|
27 |
from modules.utils import save_image
|
28 |
from modules.logger_util import get_logger
|
29 |
|
@@ -53,6 +54,7 @@ class SamInference:
|
|
53 |
self.image_predictor = None
|
54 |
self.video_predictor = None
|
55 |
self.video_inference_state = None
|
|
|
56 |
|
57 |
def load_model(self,
|
58 |
model_type: Optional[str] = None,
|
@@ -79,7 +81,6 @@ class SamInference:
|
|
79 |
)
|
80 |
except Exception as e:
|
81 |
logger.exception("Error while loading SAM2 model for video predictor")
|
82 |
-
raise f"Error while loading SAM2 model for video predictor!: {e}"
|
83 |
|
84 |
try:
|
85 |
self.model = build_sam2(
|
@@ -89,7 +90,6 @@ class SamInference:
|
|
89 |
)
|
90 |
except Exception as e:
|
91 |
logger.exception("Error while loading SAM2 model")
|
92 |
-
raise f"Error while loading SAM2 model!: {e}"
|
93 |
|
94 |
def init_video_inference_state(self,
|
95 |
vid_input: str,
|
@@ -101,11 +101,18 @@ class SamInference:
|
|
101 |
self.current_model_type = model_type
|
102 |
self.load_model(model_type=model_type, load_video_predictor=True)
|
103 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
104 |
if self.video_inference_state is not None:
|
105 |
self.video_predictor.reset_state(self.video_inference_state)
|
106 |
self.video_inference_state = None
|
107 |
|
108 |
-
self.video_inference_state = self.video_predictor.init_state(video_path=
|
109 |
|
110 |
def generate_mask(self,
|
111 |
image: np.ndarray,
|
@@ -147,7 +154,6 @@ class SamInference:
|
|
147 |
)
|
148 |
except Exception as e:
|
149 |
logger.exception(f"Error while predicting image with prompt: {str(e)}")
|
150 |
-
raise RuntimeError(f"Error while predicting image with prompt: {str(e)}") from e
|
151 |
return masks, scores, logits
|
152 |
|
153 |
def add_prediction_to_frame(self,
|
@@ -160,7 +166,6 @@ class SamInference:
|
|
160 |
if (self.video_predictor is None or
|
161 |
inference_state is None and self.video_inference_state is None):
|
162 |
logger.exception("Error while predicting frame from video, load video predictor first")
|
163 |
-
raise f"Error while predicting frame from video"
|
164 |
|
165 |
if inference_state is None:
|
166 |
inference_state = self.video_inference_state
|
@@ -184,7 +189,6 @@ class SamInference:
|
|
184 |
inference_state: Optional[Dict] = None,):
|
185 |
if inference_state is None and self.video_inference_state is None:
|
186 |
logger.exception("Error while propagating in video, load video predictor first")
|
187 |
-
raise f"Error while propagating in video"
|
188 |
|
189 |
if inference_state is None:
|
190 |
inference_state = self.video_inference_state
|
@@ -196,7 +200,6 @@ class SamInference:
|
|
196 |
inference_state=inference_state,
|
197 |
start_frame_idx=0
|
198 |
)
|
199 |
-
cached_images = inference_state["images"]
|
200 |
images = get_frames_from_dir(vid_dir=TEMP_DIR, as_numpy=True)
|
201 |
|
202 |
with torch.autocast(device_type=self.device, dtype=torch.float16):
|
@@ -208,7 +211,6 @@ class SamInference:
|
|
208 |
}
|
209 |
except Exception as e:
|
210 |
logger.exception(f"Error while propagating in video: {str(e)}")
|
211 |
-
raise RuntimeError(f"Failed to propagate in video: {str(e)}") from e
|
212 |
|
213 |
return video_segments
|
214 |
|
@@ -255,12 +257,13 @@ class SamInference:
|
|
255 |
|
256 |
return image
|
257 |
|
258 |
-
def
|
259 |
-
|
260 |
-
|
261 |
-
|
262 |
-
|
263 |
-
|
|
|
264 |
if self.video_predictor is None or self.video_inference_state is None:
|
265 |
logger.exception("Error while adding filter to preview, load video predictor first")
|
266 |
raise f"Error while adding filter to preview"
|
@@ -299,6 +302,14 @@ class SamInference:
|
|
299 |
|
300 |
save_image(image=filtered_image, output_dir=TEMP_OUT_DIR)
|
301 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
302 |
def divide_layer(self,
|
303 |
image_input: np.ndarray,
|
304 |
image_prompt_input_data: Dict,
|
|
|
23 |
create_mask_pixelized_image,
|
24 |
create_solid_color_mask_image
|
25 |
)
|
26 |
+
from modules.video_utils import (get_frames_from_dir, create_video_from_frames, get_video_info, extract_frames,
|
27 |
+
extract_sound, clean_temp_dir)
|
28 |
from modules.utils import save_image
|
29 |
from modules.logger_util import get_logger
|
30 |
|
|
|
54 |
self.image_predictor = None
|
55 |
self.video_predictor = None
|
56 |
self.video_inference_state = None
|
57 |
+
self.video_info = None
|
58 |
|
59 |
def load_model(self,
|
60 |
model_type: Optional[str] = None,
|
|
|
81 |
)
|
82 |
except Exception as e:
|
83 |
logger.exception("Error while loading SAM2 model for video predictor")
|
|
|
84 |
|
85 |
try:
|
86 |
self.model = build_sam2(
|
|
|
90 |
)
|
91 |
except Exception as e:
|
92 |
logger.exception("Error while loading SAM2 model")
|
|
|
93 |
|
94 |
def init_video_inference_state(self,
|
95 |
vid_input: str,
|
|
|
101 |
self.current_model_type = model_type
|
102 |
self.load_model(model_type=model_type, load_video_predictor=True)
|
103 |
|
104 |
+
self.video_info = get_video_info(vid_input)
|
105 |
+
frames_temp_dir = TEMP_DIR
|
106 |
+
clean_temp_dir(frames_temp_dir)
|
107 |
+
extract_frames(vid_input, frames_temp_dir)
|
108 |
+
if self.video_info.has_sound:
|
109 |
+
extract_sound(vid_input, frames_temp_dir)
|
110 |
+
|
111 |
if self.video_inference_state is not None:
|
112 |
self.video_predictor.reset_state(self.video_inference_state)
|
113 |
self.video_inference_state = None
|
114 |
|
115 |
+
self.video_inference_state = self.video_predictor.init_state(video_path=frames_temp_dir)
|
116 |
|
117 |
def generate_mask(self,
|
118 |
image: np.ndarray,
|
|
|
154 |
)
|
155 |
except Exception as e:
|
156 |
logger.exception(f"Error while predicting image with prompt: {str(e)}")
|
|
|
157 |
return masks, scores, logits
|
158 |
|
159 |
def add_prediction_to_frame(self,
|
|
|
166 |
if (self.video_predictor is None or
|
167 |
inference_state is None and self.video_inference_state is None):
|
168 |
logger.exception("Error while predicting frame from video, load video predictor first")
|
|
|
169 |
|
170 |
if inference_state is None:
|
171 |
inference_state = self.video_inference_state
|
|
|
189 |
inference_state: Optional[Dict] = None,):
|
190 |
if inference_state is None and self.video_inference_state is None:
|
191 |
logger.exception("Error while propagating in video, load video predictor first")
|
|
|
192 |
|
193 |
if inference_state is None:
|
194 |
inference_state = self.video_inference_state
|
|
|
200 |
inference_state=inference_state,
|
201 |
start_frame_idx=0
|
202 |
)
|
|
|
203 |
images = get_frames_from_dir(vid_dir=TEMP_DIR, as_numpy=True)
|
204 |
|
205 |
with torch.autocast(device_type=self.device, dtype=torch.float16):
|
|
|
211 |
}
|
212 |
except Exception as e:
|
213 |
logger.exception(f"Error while propagating in video: {str(e)}")
|
|
|
214 |
|
215 |
return video_segments
|
216 |
|
|
|
257 |
|
258 |
return image
|
259 |
|
260 |
+
def create_filtered_video(self,
|
261 |
+
image_prompt_input_data: Dict,
|
262 |
+
filter_mode: str,
|
263 |
+
frame_idx: int,
|
264 |
+
pixel_size: Optional[int] = None,
|
265 |
+
color_hex: Optional[str] = None
|
266 |
+
):
|
267 |
if self.video_predictor is None or self.video_inference_state is None:
|
268 |
logger.exception("Error while adding filter to preview, load video predictor first")
|
269 |
raise f"Error while adding filter to preview"
|
|
|
302 |
|
303 |
save_image(image=filtered_image, output_dir=TEMP_OUT_DIR)
|
304 |
|
305 |
+
out_video = create_video_from_frames(
|
306 |
+
frames_dir=TEMP_DIR,
|
307 |
+
frame_rate=self.video_info.frame_rate,
|
308 |
+
output_dir=self.output_dir,
|
309 |
+
)
|
310 |
+
|
311 |
+
return out_video, out_video
|
312 |
+
|
313 |
def divide_layer(self,
|
314 |
image_input: np.ndarray,
|
315 |
image_prompt_input_data: Dict,
|