Recurrent Metric Networks and Batch Multiple Hypothesis for Multi-Object Tracking

	Recurrent Metric Networks and Batch Multiple Hypothesis for Multi-Object Tracking
	LONGTAO CHEN; XIAOJIANG PENG; MINGWU REN
刊名	IEEE ACCESS
	2018
文献子类	期刊论文
英文摘要	Multi-object tracking aims to recover object trajectories given multiple detections in video frames. Object feature extraction and similarity metric are two keys to reliably associate trajectories. In this paper, we propose the recurrent metric network (RMNet), a CNN-RNN based similarity metric framework for multi-object tracking. Given a reference object, the RMNet takes as input random positive andnegativedetectionsandoutputssimilarityscoresovertime.TheRMNethandlesthelong-termtemporal object variations and false object detections by its long-short memory units. With the scores from RMNet, we introduce a batch multiple hypothesis (BMH) strategy, a simple yet efﬁcient data association method for batch multi-object tracking. BMH generates a hypothesis tree for each object with a dual-threshold hypothesisgenerationapproach,andthenselectsthebestbranch(orhypothesis)foreachobjectasthebatch tracking result. Specially, we model hypothesis selection as a 0-1 programming problem and introduce a rewardfunctiontore-ﬁndobjectsincaseofmissingdetection.WeevaluateourRMNetandBMHstrategyon severalpopulardatasets:2DMOT2015,PETS2009,TUD,andKITTI.Weachieveperformancecomparable or superior to these of the state-of-the-art methods.
URL标识	查看原文
语种	英语
内容类型	期刊论文
源URL	[http://ir.siat.ac.cn:8080/handle/172644/13472]
专题	深圳先进技术研究院_集成所
推荐引用方式 GB/T 7714	LONGTAO CHEN,XIAOJIANG PENG,MINGWU REN. Recurrent Metric Networks and Batch Multiple Hypothesis for Multi-Object Tracking[J]. IEEE ACCESS,2018.
APA	LONGTAO CHEN,XIAOJIANG PENG,&MINGWU REN.(2018).Recurrent Metric Networks and Batch Multiple Hypothesis for Multi-Object Tracking.IEEE ACCESS.
MLA	LONGTAO CHEN,et al."Recurrent Metric Networks and Batch Multiple Hypothesis for Multi-Object Tracking".IEEE ACCESS (2018).