Localized Multiple Kernel Learning for Realistic Human Action Recognition in Videos

doi:10.1109/TCSVT.2011.2130230

	Localized Multiple Kernel Learning for Realistic Human Action Recognition in Videos
	Song, Yan 1; Zheng, Yan-Tao 2; Tang, Sheng ; Zhou, Xiangdong 3; Zhang, Yongdong ; Lin, Shouxun ; Chua, Tat-Seng 4
	2011-09-01
发表期刊	IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
ISSN	1051-8215
卷号	21 期号:9 页码:1193-1202
摘要	Realistic human action recognition in videos has been a useful yet challenging task. Video shots of same actions may present huge intra-class variations in terms of visual appearance, kinetic patterns, video shooting, and editing styles. Heterogeneous feature representations of videos pose another challenge on how to effectively handle the redundancy, complementariness and disagreement in these features. This paper proposes a localized multiple kernel learning (L-MKL) algorithm to tackle the issues above. L-MKL integrates the localized classifier ensemble learning and multiple kernel learning in a unified framework to leverage the strengths of both. The basis of L-MKL is to build multiple kernel classifiers on diverse features at subspace localities of heterogeneous representations. L-MKL integrates the discriminability of complementary features locally and enables localized MKL classifiers to deliver better performance in its own region of expertise. Specifically, L-MKL develops a locality gating model to partition the input space of heterogeneous representations to a set of localities of simpler data structure. Each locality then learns its localized optimal combination of Mercer kernels of heterogeneous features. Finally, the gating model coordinates the localized multiple kernel classifiers globally to perform action recognition. Experiments on two datasets show that the proposed approach delivers promising performance.
关键词	Action recognition localized classifier multiple kernel learning
DOI	10.1109/TCSVT.2011.2130230
收录类别	SCI
语种	英语
资助项目	National Basic Research Program of China (973 Program)[2007CB311105] ; National Nature Science Foundation of China[60873165] ; Beijing Municipal Education Commission
WOS研究方向	Engineering
WOS类目	Engineering, Electrical & Electronic
WOS记录号	WOS:000294669900002
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
引用统计	被引频次：24[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/2XEOYT63/12980
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Song, Yan
作者单位	1.Chinese Acad Sci, Lab Adv Comp Res, Inst Comp Technol, Beijing 100190, Peoples R China 2.ASTAR, Inst Infocomm Res, Singapore, Singapore 3.Fudan Univ, Sch Comp Sci & Technol, Shanghai 200433, Peoples R China 4.Natl Univ Singapore, Sch Comp, Singapore 117548, Singapore
推荐引用方式 GB/T 7714	Song, Yan,Zheng, Yan-Tao,Tang, Sheng,et al. Localized Multiple Kernel Learning for Realistic Human Action Recognition in Videos[J]. IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY,2011,21(9):1193-1202.
APA	Song, Yan.,Zheng, Yan-Tao.,Tang, Sheng.,Zhou, Xiangdong.,Zhang, Yongdong.,...&Chua, Tat-Seng.(2011).Localized Multiple Kernel Learning for Realistic Human Action Recognition in Videos.IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY,21(9),1193-1202.
MLA	Song, Yan,et al."Localized Multiple Kernel Learning for Realistic Human Action Recognition in Videos".IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY 21.9(2011):1193-1202.