app.py

# app.py - YOLOv12 deployment for Hugging Face Spaces
import gradio as gr
from ultralytics import YOLO
from PIL import Image

# Load your YOLOv12 model
model = YOLO('best.pt')

def predict(image):
    """
    Run YOLOv12 inference on the uploaded image
    """
    try:
        # Run inference
        results = model(image)
        
        # Get the annotated image with bounding boxes
        annotated_img = results[0].plot()
        
        # Convert BGR to RGB (OpenCV to PIL)
        annotated_img = Image.fromarray(annotated_img[..., ::-1])
        
        # Get detection details
        detections = results[0].boxes
        detection_text = []
        
        for box in detections:
            cls = int(box.cls[0])
            conf = float(box.conf[0])
            class_name = model.names[cls]
            detection_text.append(f"{class_name}: {conf:.3%}")
        
        if detection_text:
            summary = "\n".join(detection_text)
        else:
            summary = "No objects detected"
        
        return annotated_img, summary
    
    except Exception as e:
        return None, f"Error: {str(e)}"

# Create Gradio interface
demo = gr.Interface(
    fn=predict,
    inputs=gr.Image(type="pil", label="Upload Image"),
    outputs=[
        gr.Image(type="pil", label="Detection Results"),
        gr.Textbox(label="Detected Objects", lines=10)
    ],
    title="Fine-Tuned YOLOv12 using Optuna",
    description="Upload an image to detect objects using YOLOv12",
)

if __name__ == "__main__":
    demo.launch()