Mixed-Supervised Scene Text Detection With Expectation-Maximization Algorithm

doi:10.1109/TIP.2022.3197987

CORC > 自动化研究所 > 中国科学院自动化研究所 > 模式识别国家重点实验室 > 模式分析与学习团队

	Mixed-Supervised Scene Text Detection With Expectation-Maximization Algorithm
	Zhao, Mengbiao 1,2; Feng, Wei 1,2; Yin, Fei 1,2; Zhang, Xu-Yao 1,2; Liu, Cheng-Lin 1,2
刊名	IEEE TRANSACTIONS ON IMAGE PROCESSING
	2022
卷号	31 页码:5513-5528
关键词	Costs Annotations Training Labeling Detectors Data models Benchmark testing Mixed-supervised learning scene text detection weak supervision forms expectation-maximization algorithm
ISSN号	1057-7149
DOI	10.1109/TIP.2022.3197987
通讯作者	Liu, Cheng-Lin(liucl@nlpr.ia.ac.cn)
英文摘要	Scene text detection is an important and challenging task in computer vision. For detecting arbitrarily-shaped texts, most existing methods require heavy data labeling efforts to produce polygon-level text region labels for supervised training. In order to reduce the cost in data labeling, we study mixed-supervised arbitrarily-shaped text detection by combining various weak supervision forms (e.g., image-level tags, coarse, loose and tight bounding boxes), which are far easier to annotate. Whereas the existing weakly-supervised learning methods (such as multiple instance learning) do not promote full object coverage, to approximate the performance of fully-supervised detection, we propose an Expectation-Maximization (EM) based mixed-supervised learning framework to train scene text detector using only a small amount of polygon-level annotated data combined with a large amount of weakly annotated data. The polygon-level labels are treated as latent variables and recovered from the weak labels by the EM algorithm. A new contour-based scene text detector is also proposed to facilitate the use of weak labels in our mixed-supervised learning framework. Extensive experiments on six scene text benchmarks show that (1) using only 10% strongly annotated data and 90% weakly annotated data, our method yields comparable performance to that of fully supervised methods, (2) with 100% strongly annotated data, our method achieves state-of-the-art performance on five scene text benchmarks (CTW1500, Total-Text, ICDAR-ArT, MSRA-TD500, and C-SVT), and competitive results on the ICDAR2015 Dataset. We will make our weakly annotated datasets publicly available.
资助项目	National Key Research and Development Program[2020AAA0108003] ; National Natural Science Foundation of China (NSFC)[61733007] ; National Natural Science Foundation of China (NSFC)[61721004]
WOS关键词	LOCALIZATION
WOS研究方向	Computer Science ; Engineering
语种	英语
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
WOS记录号	WOS:000844128200003
资助机构	National Key Research and Development Program ; National Natural Science Foundation of China (NSFC)
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/49873]
专题	自动化研究所_模式识别国家重点实验室_模式分析与学习团队
通讯作者	Liu, Cheng-Lin
作者单位	1.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing 100049, Peoples R China 2.Chinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China
推荐引用方式 GB/T 7714	Zhao, Mengbiao,Feng, Wei,Yin, Fei,et al. Mixed-Supervised Scene Text Detection With Expectation-Maximization Algorithm[J]. IEEE TRANSACTIONS ON IMAGE PROCESSING,2022,31:5513-5528.
APA	Zhao, Mengbiao,Feng, Wei,Yin, Fei,Zhang, Xu-Yao,&Liu, Cheng-Lin.(2022).Mixed-Supervised Scene Text Detection With Expectation-Maximization Algorithm.IEEE TRANSACTIONS ON IMAGE PROCESSING,31,5513-5528.
MLA	Zhao, Mengbiao,et al."Mixed-Supervised Scene Text Detection With Expectation-Maximization Algorithm".IEEE TRANSACTIONS ON IMAGE PROCESSING 31(2022):5513-5528.