A Repo For Document AI

mPLUG-DocOwl: Modularized Multimodal Large Language Model for Document Understanding

Parsing-free RAG supported by VLMs

Code for the paper "PICK: Processing Key Information Extraction from Documents using Improved Graph Learning-Convolutional Networks" (ICPR 2020)

Official PyTorch implementation of LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding (ACL 2022)

Minimal sharded dataset loaders, decoders, and utils for multi-modal document, image, and text datasets.

Datasets and Evaluation Scripts for CompHRDoc

[MM'2024] PEneo, an effective algorithm for key-value pair extraction from form-like documents, designed for real-world applications.

🐊 Snappy: A vision-first document retrieval framework, using ColPali + DeepSeek OCR. Like the project? Drop me a star! ⭐

Run optical character recognition with PyTesseract from the FiftyOne App!

A monorepo containing various utility scripts, tools, and applications for development, automation, and AI-powered tasks.

将语雀知识库接入大语言模型，实现基于 RAG（检索增强生成）的智能问答系统，支持FastAPI，兼容OpenAI API与本地Ollama模型。

This small module connects Label Studio with Fonduer by creating a fonduer labeling function for gold labels from a label studio export. Documentation: https://irgroup.github.io/labelstudio-to-fonduer/

Official implementation for "SlimDoc: Lightweight Distillation of Document Transformer Models," published in the International Journal on Document Analysis and Recognition (IJDAR), 2025

Multimodal benchmark for evaluating handwritten editorial correction in printed text.

This repository includes the ReceiptVQA dataset and the Pytorch implementation of the LiGT method and other evaluated baselines.

Analysing expense reports/invoices with AWS Textract and boto3.

PDF Chatbot is an AI-driven application that lets users chat with their PDF documents. It extracts text from uploaded PDFs and uses a powerful language model to answer user queries in a context-aware manner. The chatbot is built with Python, Gradio for the web interface, PyPDF2 for PDF parsing, and Hugging Face Transformers + LangChain for natural

Extract structured menu information from images into JSON by E2E Vision-Language model fine-tuning pipeline or LLM.