A Two-Level Rectification Attention Network for Scene Text Recognition

doi:10.1109/TMM.2022.3146779

CORC > 自动化研究所 > 中国科学院自动化研究所 > 多模态人工智能系统全国重点实验室

	A Two-Level Rectification Attention Network for Scene Text Recognition
	Wu, Lintai 1,8,9; Xu, Yong 2,8,9; Hou, Junhui 1,3; Chen, C. L. Philip 4,5; Liu, Cheng-Lin 6,7
刊名	IEEE TRANSACTIONS ON MULTIMEDIA
	2023
卷号	25 页码:2404-2414
关键词	Scene text recognition text rectification spatial transformer network optical character recognition
ISSN号	1520-9210
DOI	10.1109/TMM.2022.3146779
通讯作者	Xu, Yong(yongxu@ymail.com)
英文摘要	Scene text recognition is a challenging task in the computer vision field due to the diversity of text styles and the complexity of the image backgrounds. In recent decades, numerous text rectification and recognition methods have been proposed to solve these problems. However, most of these methods rectify texts at the geometry level or pixel level. The former is limited by geometric constraints, and the latter is prone to blurring the text. In this paper, we propose a two-level rectification attention network (TRAN) to rectify and recognize texts. This network consists of two parts: a two-level rectification network (TORN) and an attention-based recognition network (ABRN). Specifically, the TORN first rectifies texts at the geometry level and then performs a pixel-level adjustment, which not only eliminates the geometric constraints but also renders clear texts. The ABRN's role is to recognize text in the rectified images. To improve the feature extraction ability of our model, we design a new channel-wise and kernel-wise attention unit, which enables the network to handle significant variations of character size and channel interdependencies. Furthermore, we propose a skip training strategy to make our model converge smoothly. We conduct experiments on various benchmarks, including regular and irregular datasets. The experimental results show that our method achieves a state-of-the-art performance.
资助项目	National Nature Science Foundation of China[61876051] ; Shenzhen Key Laboratory of Visual Object Detection and Recognition[ZDSYS20190902093015527]
WOS关键词	EFFICIENT
WOS研究方向	Computer Science ; Telecommunications
语种	英语
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
WOS记录号	WOS:001007432100062
资助机构	National Nature Science Foundation of China ; Shenzhen Key Laboratory of Visual Object Detection and Recognition
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/53681]
专题	多模态人工智能系统全国重点实验室
通讯作者	Xu, Yong
作者单位	1.City Univ Hong Kong, Dept Comp Sci, Hong Kong, Peoples R China 2.Peng Cheng Lab, Shenzhen 518055, Guangdong, Peoples R China 3.City Univ Hong Kong, Shenzhen Res Inst, Hong Kong, Peoples R China 4.South China Univ Technol, Sch Comp Sci & Engn, Guangzhou 510006, Peoples R China 5.Pazhou Lab, Guangzhou 510335, Peoples R China 6.Chinese Acad Sci, Inst Automat, NLPR, Beijing 100190, Peoples R China 7.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing 100049, Peoples R China 8.Shenzhen Key Lab Visual Object Detect & Recognit, Shenzhen 518055, Guangdong, Peoples R China 9.Harbin Inst Technol, Biocomp Res Ctr, Shenzhen 518055, Guangdong, Peoples R China
推荐引用方式 GB/T 7714	Wu, Lintai,Xu, Yong,Hou, Junhui,et al. A Two-Level Rectification Attention Network for Scene Text Recognition[J]. IEEE TRANSACTIONS ON MULTIMEDIA,2023,25:2404-2414.
APA	Wu, Lintai,Xu, Yong,Hou, Junhui,Chen, C. L. Philip,&Liu, Cheng-Lin.(2023).A Two-Level Rectification Attention Network for Scene Text Recognition.IEEE TRANSACTIONS ON MULTIMEDIA,25,2404-2414.
MLA	Wu, Lintai,et al."A Two-Level Rectification Attention Network for Scene Text Recognition".IEEE TRANSACTIONS ON MULTIMEDIA 25(2023):2404-2414.