Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
@@ -11,9 +11,8 @@ model = DetrForObjectDetection.from_pretrained("facebook/detr-resnet-50", config
|
|
11 |
image_processor = DetrImageProcessor.from_pretrained("facebook/detr-resnet-50")
|
12 |
|
13 |
def load_model(threshold):
|
14 |
-
#
|
15 |
-
|
16 |
-
config = DetrConfig.from_pretrained("facebook/detr-resnet-50", num_labels=91, threshold=threshold)
|
17 |
model = DetrForObjectDetection.from_pretrained("facebook/detr-resnet-50", config=config)
|
18 |
image_processor = DetrImageProcessor.from_pretrained("facebook/detr-resnet-50")
|
19 |
return pipeline(task='object-detection', model=model, image_processor=image_processor)
|
@@ -27,6 +26,7 @@ def draw_detections(image, detections):
|
|
27 |
# Convert RGB to BGR for OpenCV
|
28 |
np_image = cv2.cvtColor(np_image, cv2.COLOR_RGB2BGR)
|
29 |
|
|
|
30 |
for detection in detections:
|
31 |
score = detection['score']
|
32 |
label = detection['label']
|
@@ -36,10 +36,10 @@ def draw_detections(image, detections):
|
|
36 |
x_max = box['xmax']
|
37 |
y_max = box['ymax']
|
38 |
|
39 |
-
#
|
40 |
cv2.rectangle(np_image, (x_min, y_min), (x_max, y_max), (0, 255, 0), 2)
|
41 |
label_text = f'{label} {score:.2f}'
|
42 |
-
cv2.putText(np_image, label_text, (x_min, y_min - 10), cv2.FONT_HERSHEY_SIMPLEX,
|
43 |
|
44 |
# Convert BGR to RGB for displaying
|
45 |
final_image = cv2.cvtColor(np_image, cv2.COLOR_BGR2RGB)
|
@@ -48,30 +48,39 @@ def draw_detections(image, detections):
|
|
48 |
|
49 |
def get_pipeline_prediction(threshold, pil_image):
|
50 |
global od_pipe
|
51 |
-
if od_pipe.config.threshold != threshold:
|
52 |
-
od_pipe = load_model(threshold)
|
53 |
try:
|
54 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
55 |
pipeline_output = od_pipe(pil_image)
|
56 |
processed_image = draw_detections(pil_image, pipeline_output)
|
57 |
return processed_image, pipeline_output
|
58 |
except Exception as e:
|
59 |
-
|
60 |
-
|
|
|
61 |
|
62 |
-
#
|
63 |
with gr.Blocks() as demo:
|
64 |
with gr.Row():
|
65 |
with gr.Column():
|
66 |
inp_image = gr.Image(label="Input image")
|
67 |
-
slider = gr.Slider(minimum=0, maximum=1, step=0.05, label="
|
68 |
-
gr.Markdown("Adjust the slider to change
|
69 |
btn_run = gr.Button('Run Detection')
|
70 |
with gr.Column():
|
71 |
with gr.Tab("Annotated Image"):
|
72 |
out_image = gr.Image()
|
73 |
with gr.Tab("Detection Results"):
|
74 |
out_json = gr.JSON()
|
|
|
75 |
btn_run.click(get_pipeline_prediction, inputs=[slider, inp_image], outputs=[out_image, out_json])
|
76 |
|
77 |
demo.launch()
|
|
|
11 |
image_processor = DetrImageProcessor.from_pretrained("facebook/detr-resnet-50")
|
12 |
|
13 |
def load_model(threshold):
|
14 |
+
# Reinitialize the model with the desired detection threshold
|
15 |
+
config = DetrConfig.from_pretrained("facebook/detr-resnet-50", threshold=threshold)
|
|
|
16 |
model = DetrForObjectDetection.from_pretrained("facebook/detr-resnet-50", config=config)
|
17 |
image_processor = DetrImageProcessor.from_pretrained("facebook/detr-resnet-50")
|
18 |
return pipeline(task='object-detection', model=model, image_processor=image_processor)
|
|
|
26 |
# Convert RGB to BGR for OpenCV
|
27 |
np_image = cv2.cvtColor(np_image, cv2.COLOR_RGB2BGR)
|
28 |
|
29 |
+
# Draw detections
|
30 |
for detection in detections:
|
31 |
score = detection['score']
|
32 |
label = detection['label']
|
|
|
36 |
x_max = box['xmax']
|
37 |
y_max = box['ymax']
|
38 |
|
39 |
+
# Increase font size for better visibility
|
40 |
cv2.rectangle(np_image, (x_min, y_min), (x_max, y_max), (0, 255, 0), 2)
|
41 |
label_text = f'{label} {score:.2f}'
|
42 |
+
cv2.putText(np_image, label_text, (x_min, y_min - 10), cv2.FONT_HERSHEY_SIMPLEX, 1.5, (255, 255, 255), 4)
|
43 |
|
44 |
# Convert BGR to RGB for displaying
|
45 |
final_image = cv2.cvtColor(np_image, cv2.COLOR_BGR2RGB)
|
|
|
48 |
|
49 |
def get_pipeline_prediction(threshold, pil_image):
|
50 |
global od_pipe
|
|
|
|
|
51 |
try:
|
52 |
+
# Check if the model threshold needs adjusting
|
53 |
+
if od_pipe.config.threshold != threshold:
|
54 |
+
od_pipe = load_model(threshold)
|
55 |
+
print("Model reloaded with new threshold:", threshold)
|
56 |
+
|
57 |
+
# Ensure input is a PIL image
|
58 |
+
if not isinstance(pil_image, Image.Image):
|
59 |
+
pil_image = Image.fromarray(np.array(pil_image).astype('uint8'), 'RGB')
|
60 |
+
|
61 |
+
# Run detection and return annotated image and results
|
62 |
pipeline_output = od_pipe(pil_image)
|
63 |
processed_image = draw_detections(pil_image, pipeline_output)
|
64 |
return processed_image, pipeline_output
|
65 |
except Exception as e:
|
66 |
+
error_message = f"An error occurred: {str(e)}"
|
67 |
+
print(error_message)
|
68 |
+
return pil_image, {"error": error_message}
|
69 |
|
70 |
+
# Gradio interface
|
71 |
with gr.Blocks() as demo:
|
72 |
with gr.Row():
|
73 |
with gr.Column():
|
74 |
inp_image = gr.Image(label="Input image")
|
75 |
+
slider = gr.Slider(minimum=0, maximum=1, step=0.05, label="Detection Sensitivity", value=0.5)
|
76 |
+
gr.Markdown("Adjust the slider to change detection sensitivity.")
|
77 |
btn_run = gr.Button('Run Detection')
|
78 |
with gr.Column():
|
79 |
with gr.Tab("Annotated Image"):
|
80 |
out_image = gr.Image()
|
81 |
with gr.Tab("Detection Results"):
|
82 |
out_json = gr.JSON()
|
83 |
+
|
84 |
btn_run.click(get_pipeline_prediction, inputs=[slider, inp_image], outputs=[out_image, out_json])
|
85 |
|
86 |
demo.launch()
|