app.py

import gradio as gr
from ultralytics import YOLO
import cv2
import numpy as np
import os
import requests
import torch
import huggingface_hub
from accelerate import Accelerator
from huggingface_hub import notebook_login  # Added this for HF login
from huggingface_hub.utils import HfHubHTTPError  # Added this to catch HF login errors
# Initialize Hugging Face Hub login
notebook_login()
# Initialize Accelerator
accelerator = Accelerator()


# Load the model file
model_path = "yolov8x-doclaynet-epoch64-imgsz640-initiallr1e-4-finallr1e-5.pt"
if not os.path.exists(model_path):
    # Download the model file if it doesn't exist
    model_url = "https://huggingface.co/DILHTWD/documentlayoutsegmentation_YOLOv8_ondoclaynet/resolve/main/yolov8x-doclaynet-epoch64-imgsz640-initiallr1e-4-finallr1e-5.pt"
    try:
        response = requests.get(model_url)
        with open(model_path, "wb") as f:
            f.write(response.content)
    except HfHubHTTPError as e:
        if e.response.status_code == 401:
            print("Authentication error. Please login to Hugging Face Hub.")
        else:
            raise e
# Load the document segmentation model
docseg_model = YOLO(model_path) 


docseg_model = accelerator.prepare(docseg_model)

def process_image(image):
    try:
        # Convert image to the format YOLO model expects
        image = cv2.cvtColor(np.array(image), cv2.COLOR_RGB2BGR)

        # Move image to accelerator
        image = torch.from_numpy(image).to(accelerator.device)

        results = docseg_model.predict(image)
        result = results[0]  # Get the first (and usually only) result
        
        # Extract annotated image from results
        annotated_img = result.plot() 
        annotated_img = cv2.cvtColor(annotated_img, cv2.COLOR_BGR2RGB)

        # Prepare detected areas and labels as text output
        detected_areas_labels = "\n".join(
            [f"{box.label.upper()}: {box.conf:.2f}" for box in result.boxes]
        )
    except Exception as e:
        return None, f"Error during processing: {e}"  # Error handling

    return annotated_img, detected_areas_labels

# The rest of the code remains the same (Gradio interface)