Official implementation of URaG: Unified Retrieval and Generation in Multimodal LLMs for Efficient Long Document Understanding (AAAI 2026 Oral).

A Comprehensive Benchmark for Chinese Long Historical Document Understanding

[AAAI 2025] DocKylin: A Large Multimodal Model for Visual Document Understanding with Efficient Visual Slimming

The official implement of CTRNet++.

WenMind benchmark.

Official implementation of UPOCR: Towards unified pixel-level OCR interface (ICML 2024)

Evaluation of the Optical Character Recognition (OCR) capabilities of GPT-4V(ision)