Update app.py

app.py

@@ -9,34 +9,57 @@ processor = DetrImageProcessor.from_pretrained("facebook/detr-resnet-50", revisi
 model = DetrForObjectDetection.from_pretrained("facebook/detr-resnet-50", revision="no_timm")
 model.eval()
 
+# Function for image object detection
+def image_object_detection(image_pil):
+    # Process the image with the DETR model
+    inputs = processor(images=image_pil, return_tensors="pt")
+    outputs = model(**inputs)
+
+    # Convert the image to numpy array for drawing bounding boxes
+    image_np = cv2.cvtColor(cv2.cvtColor(cv2.cvtColor(np.array(image_pil), cv2.COLOR_RGB2BGR), cv2.COLOR_BGR2RGB), cv2.COLOR_RGB2BGR)
+
+    # convert outputs (bounding boxes and class logits) to COCO API
+    # let's only keep detections with score > 0.9
+    target_sizes = torch.tensor([image_pil.size[::-1]])
+    results = processor.post_process_object_detection(outputs, target_sizes=target_sizes, threshold=0.9)[0]
+
+    # Draw bounding boxes on the image
+    for score, label, box in zip(results["scores"], results["labels"], results["boxes"]):
+        box = [int(round(i)) for i in box.tolist()]
+        cv2.rectangle(image_np, (box[0], box[1]), (box[2], box[3]), (0, 255, 0), 2)
+        label_text = f"{model.config.id2label[label.item()]}: {round(score.item(), 3)}"
+        cv2.putText(image_np, label_text, (box[0], box[1] - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 2)
+
+    return image_np
+
 # Function for live object detection from the camera
 def live_object_detection(image_pil):
-    # Convert the frame to PIL Image
-    frame_pil = Image.fromarray(cv2.cvtColor(image_pil, cv2.COLOR_BGR2RGB))
-
     # Process the frame with the DETR model
-    inputs = processor(images=frame_pil, return_tensors="pt")
+    inputs = processor(images=image_pil, return_tensors="pt")
     outputs = model(**inputs)
 
     # convert outputs (bounding boxes and class logits) to COCO API
     # let's only keep detections with score > 0.9
-    target_sizes = torch.tensor([frame_pil.size[::-1]])
+    target_sizes = torch.tensor([image_pil.size[::-1]])
     results = processor.post_process_object_detection(outputs, target_sizes=target_sizes, threshold=0.9)[0]
 
-    # Draw bounding boxes on the frame
+    # Draw bounding boxes on the image
     for score, label, box in zip(results["scores"], results["labels"], results["boxes"]):
         box = [int(round(i)) for i in box.tolist()]
         cv2.rectangle(image_pil, (box[0], box[1]), (box[2], box[3]), (0, 255, 0), 2)
-        cv2.putText(image_pil, f"{model.config.id2label[label.item()]}: {round(score.item(), 3)}",
-                    (box[0], box[1] - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 2)
+        label_text = f"{model.config.id2label[label.item()]}: {round(score.item(), 3)}"
+        cv2.putText(image_pil, label_text, (box[0], box[1] - 10), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 2)
 
     return image_pil
 
 # Define the Gradio interface
 iface = gr.Interface(
-    fn=live_object_detection,
-    inputs="image",
-    outputs="image",
+    fn=[image_object_detection, live_object_detection],
+    inputs=[
+        gr.Image(type="pil", label="Upload an image for object detection", hover=True),
+        "webcam",
+    ],
+    outputs=["image", "image"],
     live=True,
 )