TextMonkey: An OCR-Free Large Multimodal Model for Understanding Document

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Liu, Yuliang, Yang, Biao, Liu, Qiang, Li, Zhang, Ma, Zhiyin, Zhang, Shuo, Bai, Xiang
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909137649008640
author Liu, Yuliang
Yang, Biao
Liu, Qiang
Li, Zhang
Ma, Zhiyin
Zhang, Shuo
Bai, Xiang
author_facet Liu, Yuliang
Yang, Biao
Liu, Qiang
Li, Zhang
Ma, Zhiyin
Zhang, Shuo
Bai, Xiang
contents We present TextMonkey, a large multimodal model (LMM) tailored for text-centric tasks. Our approach introduces enhancement across several dimensions: By adopting Shifted Window Attention with zero-initialization, we achieve cross-window connectivity at higher input resolutions and stabilize early training; We hypothesize that images may contain redundant tokens, and by using similarity to filter out significant tokens, we can not only streamline the token length but also enhance the model's performance. Moreover, by expanding our model's capabilities to encompass text spotting and grounding, and incorporating positional information into responses, we enhance interpretability. It also learns to perform screenshot tasks through finetuning. Evaluation on 12 benchmarks shows notable improvements: 5.2% in Scene Text-Centric tasks (including STVQA, TextVQA, and OCRVQA), 6.9% in Document-Oriented tasks (such as DocVQA, InfoVQA, ChartVQA, DeepForm, Kleister Charity, and WikiTableQuestions), and 2.8% in Key Information Extraction tasks (comprising FUNSD, SROIE, and POIE). It outperforms in scene text spotting with a 10.9\% increase and sets a new standard on OCRBench, a comprehensive benchmark consisting of 29 OCR-related assessments, with a score of 561, surpassing previous open-sourced large multimodal models for document understanding. Code will be released at https://github.com/Yuliang-Liu/Monkey.
format Preprint
id arxiv_https___arxiv_org_abs_2403_04473
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle TextMonkey: An OCR-Free Large Multimodal Model for Understanding Document
Liu, Yuliang
Yang, Biao
Liu, Qiang
Li, Zhang
Ma, Zhiyin
Zhang, Shuo
Bai, Xiang
Computer Vision and Pattern Recognition
Artificial Intelligence
We present TextMonkey, a large multimodal model (LMM) tailored for text-centric tasks. Our approach introduces enhancement across several dimensions: By adopting Shifted Window Attention with zero-initialization, we achieve cross-window connectivity at higher input resolutions and stabilize early training; We hypothesize that images may contain redundant tokens, and by using similarity to filter out significant tokens, we can not only streamline the token length but also enhance the model's performance. Moreover, by expanding our model's capabilities to encompass text spotting and grounding, and incorporating positional information into responses, we enhance interpretability. It also learns to perform screenshot tasks through finetuning. Evaluation on 12 benchmarks shows notable improvements: 5.2% in Scene Text-Centric tasks (including STVQA, TextVQA, and OCRVQA), 6.9% in Document-Oriented tasks (such as DocVQA, InfoVQA, ChartVQA, DeepForm, Kleister Charity, and WikiTableQuestions), and 2.8% in Key Information Extraction tasks (comprising FUNSD, SROIE, and POIE). It outperforms in scene text spotting with a 10.9\% increase and sets a new standard on OCRBench, a comprehensive benchmark consisting of 29 OCR-related assessments, with a score of 561, surpassing previous open-sourced large multimodal models for document understanding. Code will be released at https://github.com/Yuliang-Liu/Monkey.
title TextMonkey: An OCR-Free Large Multimodal Model for Understanding Document
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2403.04473