app.py

from ultralytics import YOLO  
import torch
import cv2
import numpy as np
import gradio as gr
from PIL import Image

# Load YOLOv5 model
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
model = YOLO("yolov5s.pt")  # Load pre-trained YOLOv5s model
model.to(device)
model.eval()

# Load COCO class labels
CLASS_NAMES = model.names  # YOLOv5's built-in class names

def preprocess_image(image):
    image = Image.fromarray(image)
    image = image.convert("RGB")
    return image

def detect_objects(image):
    image = preprocess_image(image)
    results = model.predict(image)  # Run YOLOv5 inference

    # Convert results to bounding box format
    image = np.array(image)
    for result in results:
        for box, cls in zip(result.boxes.xyxy, result.boxes.cls):
            x1, y1, x2, y2 = map(int, box[:4])
            class_name = CLASS_NAMES[int(cls)]  # Get class name

            # Draw bounding box
            cv2.rectangle(image, (x1, y1), (x2, y2), (255, 0, 0), 2)

            # Put class label
            cv2.putText(image, class_name, (x1, y1 - 10), cv2.FONT_HERSHEY_SIMPLEX, 
                        0.5, (255, 0, 0), 2, cv2.LINE_AA)

    return image

# Gradio UI
iface = gr.Interface(
    fn=detect_objects,
    inputs=gr.Image(type="numpy"),
    outputs=gr.Image(type="numpy"),
    live=True,
)

iface.launch()