Create app.py

app.py

@@ -0,0 +1,73 @@
+import gradio as gr
+from PIL import Image
+from ultralytics import YOLO
+import pandas as pd
+import numpy as np
+
+# COCO class names
+COCO_CLASSES = [
+    'person','bicycle','car','motorcycle','airplane','bus','train','truck','boat',
+    'traffic light','fire hydrant','stop sign','parking meter','bench','bird','cat',
+    'dog','horse','sheep','cow','elephant','bear','zebra','giraffe','backpack',
+    'umbrella','handbag','tie','suitcase','frisbee','skis','snowboard','sports ball',
+    'kite','baseball bat','baseball glove','skateboard','surfboard','tennis racket',
+    'bottle','wine glass','cup','fork','knife','spoon','bowl','banana','apple',
+    'sandwich','orange','broccoli','carrot','hot dog','pizza','donut','cake',
+    'chair','couch','potted plant','bed','dining table','toilet','tv','laptop',
+    'mouse','remote','keyboard','cell phone','microwave','oven','toaster','sink',
+    'refrigerator','book','clock','vase','scissors','teddy bear','hair drier',
+    'toothbrush'
+]
+
+# Load YOLOv8 models
+yolo_fast = YOLO("yolov8n.pt")   # fast nano
+yolo_acc = YOLO("yolo12n.pt")    # accurate small
+
+def detect_top3_with_image(image):
+    image_rgb = image.convert("RGB")
+    
+    # YOLO fast detection
+    results_fast = yolo_fast(image_rgb)[0]
+    boxes_fast = results_fast.boxes
+    top_fast = sorted(zip(boxes_fast.cls.cpu().numpy(), boxes_fast.conf.cpu().numpy()), 
+                      key=lambda x: x[1], reverse=True)[:3]
+    fast_results = [f"{COCO_CLASSES[int(cls)]} ({conf*100:.1f}%)" for cls, conf in top_fast]
+    
+    # Convert plot to RGB
+    fast_plot = results_fast.plot()
+    fast_img = Image.fromarray(np.array(fast_plot)[:,:,::-1])  # BGR to RGB
+
+    # YOLO accurate detection
+    results_acc = yolo_acc(image_rgb)[0]
+    boxes_acc = results_acc.boxes
+    top_acc = sorted(zip(boxes_acc.cls.cpu().numpy(), boxes_acc.conf.cpu().numpy()), 
+                     key=lambda x: x[1], reverse=True)[:3]
+    acc_results = [f"{COCO_CLASSES[int(cls)]} ({conf*100:.1f}%)" for cls, conf in top_acc]
+    
+    # Convert plot to RGB
+    acc_plot = results_acc.plot()
+    acc_img = Image.fromarray(np.array(acc_plot)[:,:,::-1])  # BGR to RGB
+    
+    # Top-3 table
+    df = pd.DataFrame({
+        "Rank": [1,2,3],
+        "YOLOv8n": fast_results + [""]*(3-len(fast_results)),
+        "YOLOv12n": acc_results + [""]*(3-len(acc_results))
+    })
+    
+    return df, fast_img, acc_img
+
+iface = gr.Interface(
+    fn=detect_top3_with_image,
+    inputs=gr.Image(type="pil"),
+    outputs=[
+        gr.Dataframe(headers=["Rank","YOLOv8n","YOLOv12n"], type="pandas", label="Top-3 Detections"),
+        gr.Image(label="YOLOv8n Detection Output"),
+        gr.Image(label="YOLOv12n Detection Output")
+    ],
+    title="Image Object Detection Validator",
+    description="Upload an AI-generated image to see the top-3 detected objects and the visual detection output for nano version of both YOLOv8 and YOLOv12 models."
+)
+
+if __name__ == "__main__":
+    iface.launch()