Parametric local multiview hamming distance metric learning

doi:10.1016/j.patcog.2017.06.018

	Parametric local multiview hamming distance metric learning
	Zhai, Deming 1; Liu, Xianming 1; Chang, Hong 2; Zhen, Yi 3; Chen, Xilin 2; Guo, Maozu 4; Gao, Wen 1,5
	2018-03-01
发表期刊	PATTERN RECOGNITION
ISSN	0031-3203
卷号	75 页码:250-262
摘要	Learning an appropriate distance metric is a crucial problem in pattern recognition. To confront with the scalability issue of massive data, hamming distance on binary codes is advocated since it permits exact sub-linear kNN search and meanwhile shares the advantage of efficient storage. In this paper, we study hamming metric learning in the context of multimodal data for cross-view similarity search. We present a new method called Parametric Local Multiview Hamming metric (PLMH), which learns multiview metric based on a set of local hash functions to locally adapt to the data structure of each modality. To balance locality and computational efficiency, the hash projection matrix of each instance is parameterized, with guaranteed approximation error bound, as a linear combination of basis hash projections associated with a small set of anchor points. The weak-supervisory information (side information) provided by pair wise and triplet constraints are incorporated in a coherent way to achieve semantically effective hash codes. A local optimal conjugate gradient algorithm with orthogonal rotations is designed to learn the hash functions for each bit, and the overall hash codes are learned in a sequential manner to progressively minimize the bias. Experimental evaluations on cross-media retrieval tasks demonstrate that PLMH performs competitively against the state-of-the-art methods. (C) 2017 Elsevier Ltd. All rights reserved.
关键词	Metric learning Hamming distance Hash function learning
DOI	10.1016/j.patcog.2017.06.018
收录类别	SCI
语种	英语
资助项目	Natural Science Foundation of China[61502122] ; Natural Science Foundation of China[61672193] ; Natural Science Foundation of China[61671188] ; Natural Science Foundation of China[61571164] ; Fundamental Research Funds for the Central Universities[HIT.NSRIF.201653]
WOS研究方向	Computer Science ; Engineering
WOS类目	Computer Science, Artificial Intelligence ; Engineering, Electrical & Electronic
WOS记录号	WOS:000418971900023
出版者	ELSEVIER SCI LTD
引用统计	被引频次：21[WOS] [WOS记录] [WOS相关记录]
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/6321
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Liu, Xianming
作者单位	1.Harbin Inst Technol, Sch Comp Sci & Technol, Harbin 150001, Heilongjiang, Peoples R China 2.Chinese Acad Sci, Inst Comp Technol, Key Lab Intelligent Informat Proc, Beijing 100080, Peoples R China 3.Georgia Inst Technol, Coll Comp, Atlanta, GA 30313 USA 4.Beijing Univ Civil Engn & Architecture, Sch Elect & Informat Engn, Beijing 100044, Peoples R China 5.Peking Univ, Sch Elect Engn & Comp Sci, Natl Engn Lab Video Technol, Beijing 100871, Peoples R China
推荐引用方式 GB/T 7714	Zhai, Deming,Liu, Xianming,Chang, Hong,et al. Parametric local multiview hamming distance metric learning[J]. PATTERN RECOGNITION,2018,75:250-262.
APA	Zhai, Deming.,Liu, Xianming.,Chang, Hong.,Zhen, Yi.,Chen, Xilin.,...&Gao, Wen.(2018).Parametric local multiview hamming distance metric learning.PATTERN RECOGNITION,75,250-262.
MLA	Zhai, Deming,et al."Parametric local multiview hamming distance metric learning".PATTERN RECOGNITION 75(2018):250-262.