Spaces:
Running
on
Zero
Running
on
Zero
Commit
·
942c318
1
Parent(s):
5b5416f
load video with cv2
Browse files
app.py
CHANGED
@@ -3,6 +3,7 @@ import gc
|
|
3 |
from copy import deepcopy
|
4 |
from typing import Optional
|
5 |
|
|
|
6 |
import gradio as gr
|
7 |
import numpy as np
|
8 |
import spaces
|
@@ -10,7 +11,6 @@ import torch
|
|
10 |
from gradio.themes import Soft
|
11 |
from PIL import Image, ImageDraw
|
12 |
|
13 |
-
# Prefer local transformers in the workspace
|
14 |
from transformers import AutoModel, Sam2VideoProcessor
|
15 |
|
16 |
|
@@ -32,56 +32,25 @@ def try_load_video_frames(video_path_or_url: str) -> tuple[list[Image.Image], di
|
|
32 |
"""Load video frames as PIL Images using transformers.video_utils if available,
|
33 |
otherwise fall back to OpenCV. Returns (frames, info).
|
34 |
"""
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
-
|
41 |
-
|
42 |
-
|
43 |
-
|
44 |
-
|
45 |
-
|
46 |
-
|
47 |
-
|
48 |
-
|
49 |
-
|
50 |
-
|
51 |
-
|
52 |
-
|
53 |
-
|
54 |
-
cap.release()
|
55 |
-
if fps_val and fps_val > 0:
|
56 |
-
info["fps"] = float(fps_val)
|
57 |
-
except Exception as e:
|
58 |
-
print(f"Failed to render video with cv2: {e}")
|
59 |
-
pass
|
60 |
-
return pil_frames, info
|
61 |
-
except Exception as e:
|
62 |
-
print(f"Failed to load video with transformers.video_utils: {e}")
|
63 |
-
# Fallback to OpenCV
|
64 |
-
try:
|
65 |
-
import cv2 # type: ignore
|
66 |
-
|
67 |
-
cap = cv2.VideoCapture(video_path_or_url)
|
68 |
-
frames = []
|
69 |
-
while cap.isOpened():
|
70 |
-
ret, frame = cap.read()
|
71 |
-
if not ret:
|
72 |
-
break
|
73 |
-
frame_rgb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
|
74 |
-
frames.append(Image.fromarray(frame_rgb))
|
75 |
-
# Gather fps if available
|
76 |
-
fps_val = cap.get(cv2.CAP_PROP_FPS)
|
77 |
-
cap.release()
|
78 |
-
info = {
|
79 |
-
"num_frames": len(frames),
|
80 |
-
"fps": float(fps_val) if fps_val and fps_val > 0 else None,
|
81 |
-
}
|
82 |
-
return frames, info
|
83 |
-
except Exception as e:
|
84 |
-
raise RuntimeError(f"Failed to load video: {e}")
|
85 |
|
86 |
|
87 |
def overlay_masks_on_frame(
|
@@ -196,14 +165,12 @@ def load_model_if_needed(GLOBAL_STATE: gr.State) -> tuple[AutoModel, Sam2VideoPr
|
|
196 |
GLOBAL_STATE.dtype = dtype
|
197 |
GLOBAL_STATE.model_repo_id = desired_repo
|
198 |
|
199 |
-
return model, processor, device, dtype
|
200 |
-
|
201 |
|
202 |
def ensure_session_for_current_model(GLOBAL_STATE: gr.State) -> None:
|
203 |
"""Ensure the model/processor match the selected repo and inference_session exists.
|
204 |
If a video is already loaded, re-initialize the inference session when needed.
|
205 |
"""
|
206 |
-
|
207 |
desired_repo = _model_repo_from_key(GLOBAL_STATE.model_repo_key)
|
208 |
if GLOBAL_STATE.inference_session is None or GLOBAL_STATE.session_repo_id != desired_repo:
|
209 |
if GLOBAL_STATE.video_frames:
|
@@ -213,10 +180,10 @@ def ensure_session_for_current_model(GLOBAL_STATE: gr.State) -> None:
|
|
213 |
GLOBAL_STATE.boxes_by_frame_obj.clear()
|
214 |
GLOBAL_STATE.composited_frames.clear()
|
215 |
GLOBAL_STATE.inference_session = None
|
216 |
-
GLOBAL_STATE.inference_session = processor.init_video_session(
|
217 |
-
inference_device=device,
|
218 |
video_storage_device="cpu",
|
219 |
-
dtype=dtype,
|
220 |
)
|
221 |
GLOBAL_STATE.session_repo_id = desired_repo
|
222 |
|
@@ -229,7 +196,7 @@ def init_video_session(GLOBAL_STATE: gr.State, video: str | dict) -> tuple[AppSt
|
|
229 |
GLOBAL_STATE.masks_by_frame = {}
|
230 |
GLOBAL_STATE.color_by_obj = {}
|
231 |
|
232 |
-
|
233 |
|
234 |
# Gradio Video may provide a dict with 'name' or a direct file path
|
235 |
video_path: Optional[str] = None
|
@@ -261,10 +228,10 @@ def init_video_session(GLOBAL_STATE: gr.State, video: str | dict) -> tuple[AppSt
|
|
261 |
# Try to capture original FPS if provided by loader
|
262 |
GLOBAL_STATE.video_fps = float(fps_in)
|
263 |
# Initialize session
|
264 |
-
inference_session = processor.init_video_session(
|
265 |
-
inference_device=device,
|
266 |
video_storage_device="cpu",
|
267 |
-
dtype=dtype,
|
268 |
)
|
269 |
GLOBAL_STATE.inference_session = inference_session
|
270 |
|
@@ -272,7 +239,7 @@ def init_video_session(GLOBAL_STATE: gr.State, video: str | dict) -> tuple[AppSt
|
|
272 |
max_idx = len(frames) - 1
|
273 |
status = (
|
274 |
f"Loaded {len(frames)} frames @ {GLOBAL_STATE.video_fps or 'unknown'} fps{trimmed_note}. "
|
275 |
-
f"Device: {device}, dtype: bfloat16"
|
276 |
)
|
277 |
return GLOBAL_STATE, 0, max_idx, first_frame, status
|
278 |
|
@@ -749,8 +716,6 @@ with gr.Blocks(title="SAM2 Video (Transformers) - Interactive Segmentation", the
|
|
749 |
out_path = "/tmp/sam2_playback.mp4"
|
750 |
# Prefer imageio with PyAV/ffmpeg to respect exact fps
|
751 |
try:
|
752 |
-
import cv2 # type: ignore
|
753 |
-
|
754 |
fourcc = cv2.VideoWriter_fourcc(*"mp4v")
|
755 |
writer = cv2.VideoWriter(out_path, fourcc, fps, (w, h))
|
756 |
for fr_bgr in frames_np:
|
|
|
3 |
from copy import deepcopy
|
4 |
from typing import Optional
|
5 |
|
6 |
+
import cv2
|
7 |
import gradio as gr
|
8 |
import numpy as np
|
9 |
import spaces
|
|
|
11 |
from gradio.themes import Soft
|
12 |
from PIL import Image, ImageDraw
|
13 |
|
|
|
14 |
from transformers import AutoModel, Sam2VideoProcessor
|
15 |
|
16 |
|
|
|
32 |
"""Load video frames as PIL Images using transformers.video_utils if available,
|
33 |
otherwise fall back to OpenCV. Returns (frames, info).
|
34 |
"""
|
35 |
+
|
36 |
+
cap = cv2.VideoCapture(video_path_or_url)
|
37 |
+
frames = []
|
38 |
+
print("loading video frames")
|
39 |
+
while cap.isOpened():
|
40 |
+
ret, frame = cap.read()
|
41 |
+
if not ret:
|
42 |
+
break
|
43 |
+
frame_rgb = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
|
44 |
+
frames.append(Image.fromarray(frame_rgb))
|
45 |
+
# Gather fps if available
|
46 |
+
fps_val = cap.get(cv2.CAP_PROP_FPS)
|
47 |
+
cap.release()
|
48 |
+
print("loaded video frames")
|
49 |
+
info = {
|
50 |
+
"num_frames": len(frames),
|
51 |
+
"fps": float(fps_val) if fps_val and fps_val > 0 else None,
|
52 |
+
}
|
53 |
+
return frames, info
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
54 |
|
55 |
|
56 |
def overlay_masks_on_frame(
|
|
|
165 |
GLOBAL_STATE.dtype = dtype
|
166 |
GLOBAL_STATE.model_repo_id = desired_repo
|
167 |
|
|
|
|
|
168 |
|
169 |
def ensure_session_for_current_model(GLOBAL_STATE: gr.State) -> None:
|
170 |
"""Ensure the model/processor match the selected repo and inference_session exists.
|
171 |
If a video is already loaded, re-initialize the inference session when needed.
|
172 |
"""
|
173 |
+
load_model_if_needed(GLOBAL_STATE)
|
174 |
desired_repo = _model_repo_from_key(GLOBAL_STATE.model_repo_key)
|
175 |
if GLOBAL_STATE.inference_session is None or GLOBAL_STATE.session_repo_id != desired_repo:
|
176 |
if GLOBAL_STATE.video_frames:
|
|
|
180 |
GLOBAL_STATE.boxes_by_frame_obj.clear()
|
181 |
GLOBAL_STATE.composited_frames.clear()
|
182 |
GLOBAL_STATE.inference_session = None
|
183 |
+
GLOBAL_STATE.inference_session = GLOBAL_STATE.processor.init_video_session(
|
184 |
+
inference_device=GLOBAL_STATE.device,
|
185 |
video_storage_device="cpu",
|
186 |
+
dtype=GLOBAL_STATE.dtype,
|
187 |
)
|
188 |
GLOBAL_STATE.session_repo_id = desired_repo
|
189 |
|
|
|
196 |
GLOBAL_STATE.masks_by_frame = {}
|
197 |
GLOBAL_STATE.color_by_obj = {}
|
198 |
|
199 |
+
load_model_if_needed(GLOBAL_STATE)
|
200 |
|
201 |
# Gradio Video may provide a dict with 'name' or a direct file path
|
202 |
video_path: Optional[str] = None
|
|
|
228 |
# Try to capture original FPS if provided by loader
|
229 |
GLOBAL_STATE.video_fps = float(fps_in)
|
230 |
# Initialize session
|
231 |
+
inference_session = GLOBAL_STATE.processor.init_video_session(
|
232 |
+
inference_device=GLOBAL_STATE.device,
|
233 |
video_storage_device="cpu",
|
234 |
+
dtype=GLOBAL_STATE.dtype,
|
235 |
)
|
236 |
GLOBAL_STATE.inference_session = inference_session
|
237 |
|
|
|
239 |
max_idx = len(frames) - 1
|
240 |
status = (
|
241 |
f"Loaded {len(frames)} frames @ {GLOBAL_STATE.video_fps or 'unknown'} fps{trimmed_note}. "
|
242 |
+
f"Device: {GLOBAL_STATE.device}, dtype: bfloat16"
|
243 |
)
|
244 |
return GLOBAL_STATE, 0, max_idx, first_frame, status
|
245 |
|
|
|
716 |
out_path = "/tmp/sam2_playback.mp4"
|
717 |
# Prefer imageio with PyAV/ffmpeg to respect exact fps
|
718 |
try:
|
|
|
|
|
719 |
fourcc = cv2.VideoWriter_fourcc(*"mp4v")
|
720 |
writer = cv2.VideoWriter(out_path, fourcc, fps, (w, h))
|
721 |
for fr_bgr in frames_np:
|