Divergence-Driven Consistency Training for Semi-Supervised Facial Age Estimation

doi:10.1109/TIFS.2022.3218431

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

	Divergence-Driven Consistency Training for Semi-Supervised Facial Age Estimation
	Bao, Zenghao 6,7; Tan, Zichang 4,5; Wan, Jun 3,6,7; Ma, Xibo 6,7; Guo, Guodong 2; Lei, Zhen 1,6,7
刊名	IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY
	2023
卷号	18 页码:221-232
关键词	Facial age estimation semi-supervised efficient sample selection identity consistency
ISSN号	1556-6013
DOI	10.1109/TIFS.2022.3218431
通讯作者	Wan, Jun(jun.wan@ia.ac.cn) ; Ma, Xibo(xibo.ma@nlpr.ia.ac.cn)
英文摘要	Facial age estimation has attracted considerable attention owing to its great potential in applications. However, it still falls short of reliable age estimation due to the lack of sufficient training data with accurate age labels. Using conventional semi-supervised methods to exploit unlabeled data appears to be a good solution, but it does not yield sufficient performance gains while significantly increasing training time. Therefore, to tackle these problems, we present a Divergence-driven Consistency Training (DCT) method for enhancing both efficiency and performance in this paper. Following the idea of pseudo-labeling and consistency regularization, we assign pseudo labels predicted by the teacher model to unlabeled samples and then train the student model on labeled and unlabeled samples based on consistency regularization. Based on this, we propose two main promotions. The first is the Efficient Sample Selection (ESS) strategy, which is based on the Divergence Score to select effective samples from massive unlabeled images to reduce the training time and improve efficiency. The second is Identity Consistency (IC) regularization as the additional loss function, which introduces a high dependency of aging traits on a person. Moreover, we propose Local Prediction (LP), which is a plug-and-play component, to capture local semantics. Extensive experiments on multiple age benchmark datasets, including CACD, Morph II, MIVIA, and Chalearn LAP 2015, indicate DCT outperforms the state-of-the-art approaches significantly.
资助项目	National Key Research and Development Plan[2021YFE0205700] ; External Cooperation Key Project of Chinese Academy Sciences[173211KYSB20200002] ; Chinese National Natural Science Foundation[61876179] ; Chinese National Natural Science Foundation[61961160704] ; Chinese National Natural Science Foundation[62276254] ; Chinese National Natural Science Foundation[62176256] ; Chinese National Natural Science Foundation[62106264] ; Science and Technology Development Fund of Macau[0070/2020/AMJ] ; Zhejiang Laboratory[2021KH0AB01]
WOS关键词	RECOGNITION
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:000905076700013
资助机构	National Key Research and Development Plan ; External Cooperation Key Project of Chinese Academy Sciences ; Chinese National Natural Science Foundation ; Science and Technology Development Fund of Macau ; Zhejiang Laboratory
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/51129]
专题	多模态人工智能系统全国重点实验室
通讯作者	Wan, Jun; Ma, Xibo
作者单位	1.Chinese Acad Sci, Hong Kong Inst Sci & Innovat, Ctr Artificial Intelligence & Robot, Hong Kong, Peoples R China 2.Ant Grp, Beijing 100026, Peoples R China 3.Macau Univ Sci & Technol MUST, Fac Innovat Engn, Macau, Peoples R China 4.Natl Engn Lab Deep Learning Technol & Applicat, Beijing 100101, Peoples R China 5.Baidu Res, Inst Deep Learning, Beijing 100085, Peoples R China 6.Univ Chinese Acad Sci UCAS, Sch Artificial Intelligence, Beijing 100049, Peoples R China 7.Chinese Acad Sci CASIA, Inst Automat, Natl Lab Pattern Recognit NLPR, Beijing 100190, Peoples R China
推荐引用方式 GB/T 7714	Bao, Zenghao,Tan, Zichang,Wan, Jun,et al. Divergence-Driven Consistency Training for Semi-Supervised Facial Age Estimation[J]. IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY,2023,18:221-232.
APA	Bao, Zenghao,Tan, Zichang,Wan, Jun,Ma, Xibo,Guo, Guodong,&Lei, Zhen.(2023).Divergence-Driven Consistency Training for Semi-Supervised Facial Age Estimation.IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY,18,221-232.
MLA	Bao, Zenghao,et al."Divergence-Driven Consistency Training for Semi-Supervised Facial Age Estimation".IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY 18(2023):221-232.