app.py

from transformers import VisionEncoderDecoderModel, TrOCRProcessor
import torch
from PIL import Image
import gradio as gr

# Load processor and model
model = VisionEncoderDecoderModel.from_pretrained("qualcomm/TrOCR")
processor = TrOCRProcessor.from_pretrained("qualcomm/TrOCR")

device = "cuda" if torch.cuda.is_available() else "cpu"
model.to(device)

def recognize_text(image):
    if image is None:
        return "Please upload an image."

    pixel_values = processor(images=image, return_tensors="pt").pixel_values
    pixel_values = pixel_values.to(device)

    generated_ids = model.generate(pixel_values)
    generated_text = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
    return generated_text

# Gradio interface
demo = gr.Interface(
    fn=recognize_text,
    inputs=gr.Image(type="pil"),
    outputs="text",
    title="📝 TrOCR Printed Text Recognition",
    description="Upload a printed text image and get the recognized text using Microsoft's TrOCR (Base Model).",
    allow_flagging="never"
)

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