Sorry, you need to enable JavaScript to visit this website.

facebooktwittermailshare

TEXT RECOGNITION IN IMAGES BASED ON TRANSFORMER WITH HIERARCHICAL ATTENTION

Abstract: 

Recognizing text in images has been a hot research topic in computer vision for decades due to its various application. However, the variations in text appearance in term of perspective distortion, text line curvature, text styles, etc., cause great trouble in text recognition. Inspired by the Transformer structure that achieved outstanding performance in many natural language processing related applications, we propose a new Transformer-like structure for text recognition in images, which is referred to as the Hierarchical Attention Transformer Network (HATN). The entire network can be trained end-to-end by using only images and sentence-level annotations. A new hierarchical attention mechanism is proposed to lean the character-level, word-level and sentence-level contexts more efficiently and sufficiently. Extensive experiments on seven public datasets with regular and irregular text arrangements have demonstrated that the proposed HATN can achieve accurate recognition results with high efficiency.

up
3 users have voted: yiwei zhu, Jiana Yang, CHeng Guan

Paper Details

Authors:
Yiwei Zhu, Shilin Wang, Zheng Huang and Kai Chen
Submitted On:
18 September 2019 - 10:32am
Short Link:
Type:
Poster
Event:
Paper Code:
2933
Document Year:
2019
Cite

Document Files

yiweizhu_poster.pdf

(86)

Subscribe

[1] Yiwei Zhu, Shilin Wang, Zheng Huang and Kai Chen, "TEXT RECOGNITION IN IMAGES BASED ON TRANSFORMER WITH HIERARCHICAL ATTENTION", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4673. Accessed: Sep. 23, 2020.
@article{4673-19,
url = {http://sigport.org/4673},
author = {Yiwei Zhu; Shilin Wang; Zheng Huang and Kai Chen },
publisher = {IEEE SigPort},
title = {TEXT RECOGNITION IN IMAGES BASED ON TRANSFORMER WITH HIERARCHICAL ATTENTION},
year = {2019} }
TY - EJOUR
T1 - TEXT RECOGNITION IN IMAGES BASED ON TRANSFORMER WITH HIERARCHICAL ATTENTION
AU - Yiwei Zhu; Shilin Wang; Zheng Huang and Kai Chen
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4673
ER -
Yiwei Zhu, Shilin Wang, Zheng Huang and Kai Chen. (2019). TEXT RECOGNITION IN IMAGES BASED ON TRANSFORMER WITH HIERARCHICAL ATTENTION. IEEE SigPort. http://sigport.org/4673
Yiwei Zhu, Shilin Wang, Zheng Huang and Kai Chen, 2019. TEXT RECOGNITION IN IMAGES BASED ON TRANSFORMER WITH HIERARCHICAL ATTENTION. Available at: http://sigport.org/4673.
Yiwei Zhu, Shilin Wang, Zheng Huang and Kai Chen. (2019). "TEXT RECOGNITION IN IMAGES BASED ON TRANSFORMER WITH HIERARCHICAL ATTENTION." Web.
1. Yiwei Zhu, Shilin Wang, Zheng Huang and Kai Chen. TEXT RECOGNITION IN IMAGES BASED ON TRANSFORMER WITH HIERARCHICAL ATTENTION [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4673