Deep convolutional self-paced clustering

doi:10.1007/s10489-021-02569-y

CORC > 自动化研究所 > 中国科学院自动化研究所 > 精密感知与控制研究中心 > 人工智能与机器学习

	Deep convolutional self-paced clustering
	Chen, Rui 1,2; Tang, Yongqiang 2; Tian, Lei 2,3; Zhang, Caixia 1; Zhang, Wensheng 2,3
刊名	APPLIED INTELLIGENCE
	2021-07-29
页码	15
关键词	Deep clustering Convolutional autoencoder Local structure preservation Self-paced learning
ISSN号	0924-669X
DOI	10.1007/s10489-021-02569-y
通讯作者	Tang, Yongqiang(yongqiang.tang@ia.ac.cn) ; Zhang, Caixia(zh_caixia@163.com)
英文摘要	Clustering is a crucial but challenging task in data mining and machine learning. Recently, deep clustering, which derives inspiration primarily from deep learning approaches, has achieved state-of-the-art performance in various applications and attracted considerable attention. Nevertheless, most of these approaches fail to effectively learn informative cluster-oriented features for data with spatial correlation structure, e.g., images. To tackle this problem, in this paper, we develop a deep convolutional self-paced clustering (DCSPC) method. Specifically, in the pretraining stage, we propose to utilize a convolutional autoencoder to extract a high-quality data representation that contains the spatial correlation information. Then, in the finetuning stage, a clustering loss is directly imposed on the learned features to jointly perform feature refinement and cluster assignment. We retain the decoder to avoid the feature space being distorted by the clustering loss. To stabilize the training process of the whole network, we further introduce a self-paced learning mechanism and select the most confident samples in each iteration. Through comprehensive experiments on seven popular image datasets, we demonstrate that the proposed algorithm can consistently outperform state-of-the-art rivals.
资助项目	Key-Area Research and Development Program of Guangdong Province[2019B010153002] ; National Natural Science Foundation of China[U1936206] ; National Natural Science Foundation of China[61806202] ; National Natural Science Foundation of China[61803087] ; National Natural Science Foundation of China[61803086] ; Feature Innovation Project of Guangdong Province Department of Education[2019KTSCX192] ; Guangdong Basic and Applied Basic Research Fund[2020B1515310003] ; Foshan Core Technology Research Project[1920001001367]
WOS关键词	DIMENSIONALITY
WOS研究方向	Computer Science
语种	英语
出版者	SPRINGER
WOS记录号	WOS:000679334800002
资助机构	Key-Area Research and Development Program of Guangdong Province ; National Natural Science Foundation of China ; Feature Innovation Project of Guangdong Province Department of Education ; Guangdong Basic and Applied Basic Research Fund ; Foshan Core Technology Research Project
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/45563]
专题	精密感知与控制研究中心_人工智能与机器学习
通讯作者	Tang, Yongqiang; Zhang, Caixia
作者单位	1.Foshan Univ, Dept Automat, Foshan, Peoples R China 2.Chinese Acad Sci, Inst Automat, Res Ctr Precis Sensing & Control, Beijing, Peoples R China 3.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing, Peoples R China
推荐引用方式 GB/T 7714	Chen, Rui,Tang, Yongqiang,Tian, Lei,et al. Deep convolutional self-paced clustering[J]. APPLIED INTELLIGENCE,2021:15.
APA	Chen, Rui,Tang, Yongqiang,Tian, Lei,Zhang, Caixia,&Zhang, Wensheng.(2021).Deep convolutional self-paced clustering.APPLIED INTELLIGENCE,15.
MLA	Chen, Rui,et al."Deep convolutional self-paced clustering".APPLIED INTELLIGENCE (2021):15.