OCR-APP

OCR (Optical Character Recognition) App using ColPali implementation of the new Byaldi library + Huggingface transformers for Qwen2-VL. Made By Prasoon Sharma

This project is an Optical Character Recognition (OCR) app built using Gradio and powered by deep learning models such as Qwen2-VL. The app allows users to upload an image, extract text in both Hindi and English, and optionally highlight keywords in the extracted text.

Features

• Upload an image and extract text in multiple languages (supports Hindi and English).
• Optional keyword search: Highlights specific keywords in the extracted text.
• Uses OpenCV for image preprocessing.
• Intuitive web-based UI with Gradio for easy interaction.

Setup

Requirements

• Python 3.7+
• PyTorch
• OpenCV
• Gradio
• Transformers (Hugging Face)

Install dependencies using the following command:

pip install -r requirements.txt

How to run the app

• Run the Gradio app on Huggingface
• Run on Google Collab
• Run the Python file locally

Demo and Screenshots

-Demo and Screenshots