Towards Unified Multi-granularity Text Detection with Interactive Attention

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wan, Xingyu, Zhang, Chengquan, Lyu, Pengyuan, Fan, Sen, Ni, Zihan, Yao, Kun, Ding, Errui, Wang, Jingdong
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913370274267136
author Wan, Xingyu
Zhang, Chengquan
Lyu, Pengyuan
Fan, Sen
Ni, Zihan
Yao, Kun
Ding, Errui
Wang, Jingdong
author_facet Wan, Xingyu
Zhang, Chengquan
Lyu, Pengyuan
Fan, Sen
Ni, Zihan
Yao, Kun
Ding, Errui
Wang, Jingdong
contents Existing OCR engines or document image analysis systems typically rely on training separate models for text detection in varying scenarios and granularities, leading to significant computational complexity and resource demands. In this paper, we introduce "Detect Any Text" (DAT), an advanced paradigm that seamlessly unifies scene text detection, layout analysis, and document page detection into a cohesive, end-to-end model. This design enables DAT to efficiently manage text instances at different granularities, including *word*, *line*, *paragraph* and *page*. A pivotal innovation in DAT is the across-granularity interactive attention module, which significantly enhances the representation learning of text instances at varying granularities by correlating structural information across different text queries. As a result, it enables the model to achieve mutually beneficial detection performances across multiple text granularities. Additionally, a prompt-based segmentation module refines detection outcomes for texts of arbitrary curvature and complex layouts, thereby improving DAT's accuracy and expanding its real-world applicability. Experimental results demonstrate that DAT achieves state-of-the-art performances across a variety of text-related benchmarks, including multi-oriented/arbitrarily-shaped scene text detection, document layout analysis and page detection tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2405_19765
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Towards Unified Multi-granularity Text Detection with Interactive Attention
Wan, Xingyu
Zhang, Chengquan
Lyu, Pengyuan
Fan, Sen
Ni, Zihan
Yao, Kun
Ding, Errui
Wang, Jingdong
Computer Vision and Pattern Recognition
Artificial Intelligence
Existing OCR engines or document image analysis systems typically rely on training separate models for text detection in varying scenarios and granularities, leading to significant computational complexity and resource demands. In this paper, we introduce "Detect Any Text" (DAT), an advanced paradigm that seamlessly unifies scene text detection, layout analysis, and document page detection into a cohesive, end-to-end model. This design enables DAT to efficiently manage text instances at different granularities, including *word*, *line*, *paragraph* and *page*. A pivotal innovation in DAT is the across-granularity interactive attention module, which significantly enhances the representation learning of text instances at varying granularities by correlating structural information across different text queries. As a result, it enables the model to achieve mutually beneficial detection performances across multiple text granularities. Additionally, a prompt-based segmentation module refines detection outcomes for texts of arbitrary curvature and complex layouts, thereby improving DAT's accuracy and expanding its real-world applicability. Experimental results demonstrate that DAT achieves state-of-the-art performances across a variety of text-related benchmarks, including multi-oriented/arbitrarily-shaped scene text detection, document layout analysis and page detection tasks.
title Towards Unified Multi-granularity Text Detection with Interactive Attention
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2405.19765