TextFormer: A Query-based End-to-end Text Spotter with Mixed Supervision

doi:10.1007/s11633-023-1460-6

CASIA OpenIR > 学术期刊 > Machine Intelligence Research

	TextFormer: A Query-based End-to-end Text Spotter with Mixed Supervision
	Yukun Zhai 1; Xiaoqiang Zhang 2; Xiameng Qin 2; Sanyuan Zhao 1; Xingping Dong 1; Jianbing Shen 1
发表期刊	Machine Intelligence Research
ISSN	2731-538X
	2024
卷号	21 期号:4 页码:704-717
摘要	End-to-end text spotting is a vital computer vision task that aims to integrate scene text detection and recognition into a unified framework. Typical methods heavily rely on region-of-interest (RoI) operations to extract local features and complex post-processing steps to produce final predictions. To address these limitations, we propose TextFormer, a query-based end-to-end text spotter with a transformer architecture. Specifically, using query embedding per text instance, TextFormer builds upon an image encoder and a text decoder to learn a joint semantic understanding for multitask modeling. It allows for mutual training and optimization of classification, segmentation and recognition branches, resulting in deeper feature sharing without sacrificing flexibility or simplicity. Additionally, we design an adaptive global aggregation (AGG) module to transfer global features into sequential features for reading arbitrarilyshaped texts, which overcomes the suboptimization problem of RoI operations. Furthermore, potential corpus information is utilized from weak annotations to full labels through mixed supervision, further improving text detection and end-to-end text spotting results. Extensive experiments on various bilingual (i.e., English and Chinese) benchmarks demonstrate the superiority of our method. Especially on the TDA-ReCTS dataset, TextFormer surpasses the state-of-the-art method in terms of 1-NED by 13.2%.
关键词	End-to-end text spotting arbitrarily-shaped texts transformer mixed supervision multitask modeling
DOI	10.1007/s11633-023-1460-6
引用统计
文献类型	期刊论文
条目标识符	http://ir.ia.ac.cn/handle/173211/58568
专题	学术期刊_Machine Intelligence Research
作者单位	1.School of Computer Science, Beijing Institute of Technology, Beijing 100081, China 2.Department of Computer Vision Technology, Baidu Inc., Beijing 100193, China
推荐引用方式 GB/T 7714	Yukun Zhai, Xiaoqiang Zhang, Xiameng Qin,et al. TextFormer: A Query-based End-to-end Text Spotter with Mixed Supervision[J]. Machine Intelligence Research,2024,21(4):704-717.
APA	Yukun Zhai, Xiaoqiang Zhang, Xiameng Qin, Sanyuan Zhao,Xingping Dong,& Jianbing Shen.(2024).TextFormer: A Query-based End-to-end Text Spotter with Mixed Supervision.Machine Intelligence Research,21(4),704-717.
MLA	Yukun Zhai,et al."TextFormer: A Query-based End-to-end Text Spotter with Mixed Supervision".Machine Intelligence Research 21.4(2024):704-717.

条目包含的文件		下载所有文件
文件名称/大小	文献类型	版本类型	开放类型	使用许可
MIR-2023-03-038.pdf（2312KB）	期刊论文	出版稿	开放获取	CC BY-NC-SA	浏览下载