Supervised Discrete Hashing With Relaxation

doi:10.1109/TNNLS.2016.2636870

CASIA OpenIR > 智能感知与计算研究中心

	Supervised Discrete Hashing With Relaxation
	Gui, Jie 1,2; Liu, Tongliang 3,4; Sun, Zhenan5 ; Tao, Dacheng 6; Tan, Tieniu5
发表期刊	IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
	2018-03-01
卷号	29 期号:3 页码:608-617
文章类型	Article
摘要	Data-dependent hashing has recently attracted attention due to being able to support efficient retrieval and storage of high-dimensional data, such as documents, images, and videos. In this paper, we propose a novel learning-based hashing method called "supervised discrete hashing with relaxation" (SDHR) based on "supervised discrete hashing" (SDH). SDH uses ordinary least squares regression and traditional zero-one matrix encoding of class label information as the regression target (code words), thus fixing the regression target. In SDHR, the regression target is instead optimized. The optimized regression target matrix satisfies a large margin constraint for correct classification of each example. Compared with SDH, which uses the traditional zero-one matrix, SDHR utilizes the learned regression target matrix and, therefore, more accurately measures the classification error of the regression model and is more flexible. As expected, SDHR generally outperforms SDH. Experimental results on two large-scale image data sets (CIFAR-10 and MNIST) and a large-scale and challenging face data set (FRGC) demonstrate the effectiveness and efficiency of SDHR.
关键词	Data-dependent Hashing Least Squares Regression Supervised Discrete Hashing (Sdh) Supervised Discrete Hashing With Relaxation (Sdhr)
WOS标题词	Science & Technology ; Technology
DOI	10.1109/TNNLS.2016.2636870
关键词[WOS]	LEARNING BINARY-CODES ; ITERATIVE QUANTIZATION ; PROCRUSTEAN APPROACH ; IMAGE RETRIEVAL ; RECOGNITION ; SCENE
收录类别	SCI
语种	英语
项目资助者	National Science Foundation of China(61572463 ; Open Project Program of the National Laboratory of Pattern Recognition (NLPR)(201700027) ; CCF-Tencent Open Fund ; Australian Research Council(DP-140102164 ; 61573360) ; FT-130101457 ; LE-140100061)
WOS研究方向	Computer Science ; Engineering
WOS类目	Computer Science, Artificial Intelligence ; Computer Science, Hardware & Architecture ; Computer Science, Theory & Methods ; Engineering, Electrical & Electronic
WOS记录号	WOS:000426344600009
引用统计	被引频次：66[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://ir.ia.ac.cn/handle/173211/20763
专题	智能感知与计算研究中心
通讯作者	Sun, Zhenan
作者单位	1.Chinese Acad Sci, Inst Intelligent Machines, Hefei 230031, Anhui, Peoples R China 2.Shenzhen Univ, Shenzhen Key Lab Media Secur, Shenzhen 518060, Peoples R China 3.Univ Technol Sydney, Ctr Artificial Intelligence, Sydney, NSW 2007, Australia 4.Univ Technol Sydney, Fac Engn & Informat Technol, Sydney, NSW 2007, Australia 5.Chinese Acad Sci, Ctr Res Intelligent Percept & Comp, Natl Lab Pattern Recognit, Inst Automat,CAS Ctr Excellence Brain Sci & Intel, Beijing 100190, Peoples R China 6.Univ Sydney, Sch Informat Technol, Fac Engn & Informat Technol, Sydney, NSW 2006, Australia
通讯作者单位	模式识别国家重点实验室
推荐引用方式 GB/T 7714	Gui, Jie,Liu, Tongliang,Sun, Zhenan,et al. Supervised Discrete Hashing With Relaxation[J]. IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS,2018,29(3):608-617.
APA	Gui, Jie,Liu, Tongliang,Sun, Zhenan,Tao, Dacheng,&Tan, Tieniu.(2018).Supervised Discrete Hashing With Relaxation.IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS,29(3),608-617.
MLA	Gui, Jie,et al."Supervised Discrete Hashing With Relaxation".IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 29.3(2018):608-617.