Set-label modeling and deep metric learning on person re-identification

CASIA OpenIR > 类脑智能研究中心

	Set-label modeling and deep metric learning on person re-identification
	Hao Liu; Bingpeng Ma; Lei Qin; Junbiao Pang; Chunjie Zhang; Qingming Huang
发表期刊	Neurocomputing
	2015
期号	151 页码:1283–1292
摘要	Person re-identification aimsatmatchingindividualsacrossmultiplenon-overlappingadjacentcameras. By condensingmultiplegalleryimagesofapersonasawhole,weproposeanovelmethodnamedSet- Label Model(SLM)toimprovetheperformanceofpersonre-identification underthemulti-shotsetting. Moreover,weutilizemutual-informationtomeasuretherelevancebetweenqueryimageandgallery sets. Todecreasethecomputationalcomplexity,weapplyaNaive–BayesNearest-Neighboralgorithmto approximatethemutual-informationvalue.Toovercomethelimitationsoftraditionallinearmetric learning, wefurtherdevelopadeepnon-linearmetriclearning(DeepML)approachbasedonNeighbor- hood ComponentAnalysisandDeepBeliefNetwork.Toevaluatetheeffectivenessofourproposed approaches, SLMandDeepML,wehavecarriedoutextensiveexperimentsontwochallengingdatasetsi- LIDS andETHZ.Theexperimentalresultsdemonstratethattheproposedmethodscanobtainbetter performances comparedwiththestate-of-the-artmethods.
关键词	Person Re-identification Mutual-information Metric Learning Deep Learning Neighborhoodcomponentanalysis
文献类型	期刊论文
条目标识符	http://ir.ia.ac.cn/handle/173211/15394
专题	类脑智能研究中心
推荐引用方式 GB/T 7714	Hao Liu,Bingpeng Ma,Lei Qin,et al. Set-label modeling and deep metric learning on person re-identification[J]. Neurocomputing,2015(151):1283–1292.
APA	Hao Liu,Bingpeng Ma,Lei Qin,Junbiao Pang,Chunjie Zhang,&Qingming Huang.(2015).Set-label modeling and deep metric learning on person re-identification.Neurocomputing(151),1283–1292.
MLA	Hao Liu,et al."Set-label modeling and deep metric learning on person re-identification".Neurocomputing .151(2015):1283–1292.

条目包含的文件
文件名称/大小	文献类型	版本类型	开放类型	使用许可
1-s2.0-S092523121401（2739KB）	期刊论文	作者接受稿	开放获取	CC BY-NC-SA	浏览