Micro-expression recognition using dual-view self-supervised contrastive learning with intensity perception

doi:10.1016/j.neucom.2024.129142

PSYCH OpenIR

	Micro-expression recognition using dual-view self-supervised contrastive learning with intensity perception
	Li, Jingting 1,2; Zhou, Haoliang 3,4; Qian, Yu 1,4; Dong, Zizhao 1; Wang, Su-Jing 1,2
通讯作者	Wang, Su-Jing(wangsujing@psych.ac.cn)
摘要	Micro-expressions, as indicators of true emotions, have significant applications in medical care and public safety. These expressions are characterized by their short duration, low intensity, and localized occurrence. These characteristics lead to the small sample problem in micro-expressions, making feature learning challenging and limiting the improvement of recognition performance. To address this issue, we propose a multimodal contrastive learning pre-training model based on Action Unit (AU) intensity perception. We conducted an experiment to determine the minimum threshold for recognizing facial expressions. Using this threshold, we filtered a large volume of unsupervised samples. The first stage involves unsupervised multimodal contrastive learning, where the model learns from differences in facial actions across various modalities. Subsequently, the model is trained on the micro-expression recognition task using a small amount of labeled data, overcoming the limitations of small sample sizes. Comparative experiments using the MEGC2019-CD and the multimodal dataset CAS(ME)3 datasets demonstrate the superiority of our method. Our method is available at https: //github.com/MELABIPCAS/DVSCL.git.
关键词	Micro-expression Small sample size problem Contrastive learning Self-supervised learning
	2025-02-28
语种	英语
DOI	10.1016/j.neucom.2024.129142
发表期刊	NEUROCOMPUTING
ISSN	0925-2312
卷号	619 页码:13
收录类别	SCI
资助项目	National Nat-ural Science Foundation of China[62476269] ; National Nat-ural Science Foundation of China[62276252] ; National Nat-ural Science Foundation of China[62106256] ; Youth Innovation Promo-tion Association CAS, China
出版者	ELSEVIER
WOS关键词	INFORMATION
WOS研究方向	Computer Science
WOS类目	Computer Science, Artificial Intelligence
WOS记录号	WOS:001391358100001
资助机构	National Nat-ural Science Foundation of China ; Youth Innovation Promo-tion Association CAS, China
引用统计
文献类型	期刊论文
条目标识符	https://ir.psych.ac.cn/handle/311026/48617
通讯作者	Wang, Su-Jing
作者单位	1.Chinese Acad Sci, Inst Psychol, Key Lab Behav Sci, Beijing 100101, Peoples R China 2.Univ Chinese Acad Sci, Dept Psychol, Beijing 100049, Peoples R China 3.Tianjin Univ Technol, Tianjin 300382, Peoples R China 4.Jiangsu Univ Sci & Technol, Zhenjiang 212100, Peoples R China
推荐引用方式 GB/T 7714	Li, Jingting,Zhou, Haoliang,Qian, Yu,et al. Micro-expression recognition using dual-view self-supervised contrastive learning with intensity perception[J]. NEUROCOMPUTING,2025,619:13.
APA	Li, Jingting,Zhou, Haoliang,Qian, Yu,Dong, Zizhao,&Wang, Su-Jing.(2025).Micro-expression recognition using dual-view self-supervised contrastive learning with intensity perception.NEUROCOMPUTING,619,13.
MLA	Li, Jingting,et al."Micro-expression recognition using dual-view self-supervised contrastive learning with intensity perception".NEUROCOMPUTING 619(2025):13.