Spaces:
Running
on
Zero
Running
on
Zero
Commit
·
61732db
1
Parent(s):
6a95f1f
retry
Browse files
app.py
CHANGED
@@ -2,6 +2,9 @@ import spaces
|
|
2 |
import gradio as gr
|
3 |
import cv2
|
4 |
from PIL import Image
|
|
|
|
|
|
|
5 |
|
6 |
from transformers import RTDetrForObjectDetection, RTDetrImageProcessor
|
7 |
|
@@ -16,7 +19,7 @@ def stream_object_detection(video, conf_threshold):
|
|
16 |
|
17 |
video_codec = cv2.VideoWriter_fourcc(*"x264") # type: ignore
|
18 |
fps = int(cap.get(cv2.CAP_PROP_FPS))
|
19 |
-
desired_fps = fps //
|
20 |
width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
|
21 |
height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
|
22 |
|
@@ -24,28 +27,33 @@ def stream_object_detection(video, conf_threshold):
|
|
24 |
|
25 |
n_frames = 0
|
26 |
n_chunks = 0
|
27 |
-
|
|
|
28 |
segment_file = cv2.VideoWriter(name, video_codec, fps, (width, height)) # type: ignore
|
29 |
batch = []
|
30 |
|
31 |
while iterating:
|
32 |
frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
|
33 |
-
if n_frames %
|
34 |
batch.append(frame)
|
35 |
-
if len(batch) == desired_fps:
|
36 |
inputs = image_processor(images=batch, return_tensors="pt")
|
37 |
|
|
|
|
|
38 |
with torch.no_grad():
|
39 |
outputs = model(**inputs)
|
|
|
|
|
40 |
|
41 |
boxes = image_processor.post_process_object_detection(
|
42 |
outputs,
|
43 |
-
target_sizes=torch.tensor([
|
44 |
threshold=conf_threshold)
|
45 |
|
46 |
for array, box in zip(batch, boxes):
|
47 |
-
pil_image = draw_bounding_boxes(Image.
|
48 |
-
frame =
|
49 |
# Convert RGB to BGR
|
50 |
frame = frame[:, :, ::-1].copy()
|
51 |
segment_file.write(frame)
|
@@ -54,7 +62,7 @@ def stream_object_detection(video, conf_threshold):
|
|
54 |
n_frames = 0
|
55 |
n_chunks += 1
|
56 |
yield name
|
57 |
-
name =
|
58 |
segment_file = cv2.VideoWriter(name, video_codec, fps, (width, height)) # type: ignore
|
59 |
|
60 |
iterating, frame = cap.read()
|
@@ -83,7 +91,7 @@ with gr.Blocks(css=css) as app:
|
|
83 |
""")
|
84 |
with gr.Column(elem_classes=["my-column"]):
|
85 |
with gr.Group(elem_classes=["my-group"]):
|
86 |
-
video = gr.Video(label="Video Source")
|
87 |
conf_threshold = gr.Slider(
|
88 |
label="Confidence Threshold",
|
89 |
minimum=0.0,
|
|
|
2 |
import gradio as gr
|
3 |
import cv2
|
4 |
from PIL import Image
|
5 |
+
import torch
|
6 |
+
import time
|
7 |
+
import numpy as np
|
8 |
|
9 |
from transformers import RTDetrForObjectDetection, RTDetrImageProcessor
|
10 |
|
|
|
19 |
|
20 |
video_codec = cv2.VideoWriter_fourcc(*"x264") # type: ignore
|
21 |
fps = int(cap.get(cv2.CAP_PROP_FPS))
|
22 |
+
desired_fps = fps // 5
|
23 |
width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
|
24 |
height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))
|
25 |
|
|
|
27 |
|
28 |
n_frames = 0
|
29 |
n_chunks = 0
|
30 |
+
|
31 |
+
name = f"output_{n_chunks}.ts"
|
32 |
segment_file = cv2.VideoWriter(name, video_codec, fps, (width, height)) # type: ignore
|
33 |
batch = []
|
34 |
|
35 |
while iterating:
|
36 |
frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
|
37 |
+
if n_frames % 5 == 0:
|
38 |
batch.append(frame)
|
39 |
+
if len(batch) == 2 * desired_fps:
|
40 |
inputs = image_processor(images=batch, return_tensors="pt")
|
41 |
|
42 |
+
print(f"starting batch of size {len(batch)}")
|
43 |
+
start = time.time()
|
44 |
with torch.no_grad():
|
45 |
outputs = model(**inputs)
|
46 |
+
end = time.time()
|
47 |
+
print("time taken ", end - start)
|
48 |
|
49 |
boxes = image_processor.post_process_object_detection(
|
50 |
outputs,
|
51 |
+
target_sizes=torch.tensor([frame[0].shape[:2][::-1]] * len(batch)),
|
52 |
threshold=conf_threshold)
|
53 |
|
54 |
for array, box in zip(batch, boxes):
|
55 |
+
pil_image = draw_bounding_boxes(Image.fromarray(array), box, model, conf_threshold)
|
56 |
+
frame = np.array(pil_image)
|
57 |
# Convert RGB to BGR
|
58 |
frame = frame[:, :, ::-1].copy()
|
59 |
segment_file.write(frame)
|
|
|
62 |
n_frames = 0
|
63 |
n_chunks += 1
|
64 |
yield name
|
65 |
+
name = f"output_{n_chunks}.ts"
|
66 |
segment_file = cv2.VideoWriter(name, video_codec, fps, (width, height)) # type: ignore
|
67 |
|
68 |
iterating, frame = cap.read()
|
|
|
91 |
""")
|
92 |
with gr.Column(elem_classes=["my-column"]):
|
93 |
with gr.Group(elem_classes=["my-group"]):
|
94 |
+
video = gr.Video(label="Video Source", streaming=True, autoplay=True)
|
95 |
conf_threshold = gr.Slider(
|
96 |
label="Confidence Threshold",
|
97 |
minimum=0.0,
|