Create app.py

app.py

@@ -0,0 +1,71 @@
+import gradio as gr
+import cv2
+import numpy as np
+from collections import defaultdict
+from transformers import pipeline, AutoModelForObjectDetection, AutoFeatureExtractor
+
+# Load Xenova's DETR model
+model_name = "Xenova/detr-resnet-50"
+feature_extractor = AutoFeatureExtractor.from_pretrained(model_name)
+model = AutoModelForObjectDetection.from_pretrained(model_name)
+detector = pipeline("object-detection", model=model, feature_extractor=feature_extractor)
+
+# Global counter (persists across frames)
+object_counter = defaultdict(int)
+
+def process_video(video_path):
+    cap = cv2.VideoCapture(video_path)
+    
+    while cap.isOpened():
+        ret, frame = cap.read()
+        if not ret:
+            break
+        
+        # Convert frame to RGB (required by DETR)
+        rgb_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
+        
+        # Detect objects
+        results = detector(rgb_frame)
+        
+        # Draw boxes and update counter
+        for obj in results:
+            label = obj["label"]
+            score = obj["score"]
+            box = obj["box"]
+            
+            # Update counter
+            object_counter[label] += 1
+            
+            # Draw bounding box
+            xmin, ymin, xmax, ymax = int(box["xmin"]), int(box["ymin"]), int(box["xmax"]), int(box["ymax"])
+            cv2.rectangle(frame, (xmin, ymin), (xmax, ymax), (0, 255, 0), 2)
+            cv2.putText(frame, f"{label} ({score:.2f})", (xmin, ymin-10), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0, 255, 0), 2)
+        
+        # Overlay counter on the frame
+        counter_text = "\n".join([f"{k}: {v}" for k, v in object_counter.items()])
+        cv2.putText(frame, counter_text, (10, 30), cv2.FONT_HERSHEY_SIMPLEX, 0.7, (255, 0, 0), 2)
+        
+        # Convert back to RGB for Gradio
+        output_frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
+        yield output_frame
+    
+    cap.release()
+
+# Gradio UI
+with gr.Blocks() as demo:
+    gr.Markdown("# 🎥 Video Object Detection (Xenova/detr-resnet-50)")
+    with gr.Row():
+        video_input = gr.Video(label="Upload Video")
+        video_output = gr.Image(label="Live Detection")
+    with gr.Row():
+        gr.Markdown("### Cumulative Object Counts")
+        counter_display = gr.Textbox(value="", label="Counter", interactive=False)
+    
+    video_input.change(
+        fn=process_video,
+        inputs=video_input,
+        outputs=video_output,
+        show_progress=True,
+    )
+
+demo.launch()