Bilayer Sparse Topic Model for Scene Analysis in Imbalanced Surveillance Videos

被引:13
作者
Wang, Jinqiao [1 ]
Fu, Wei [2 ]
Lu, Hanqing [1 ]
Ma, Songde [1 ]
机构
[1] Chinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China
[2] China Elect Technol Grp Corp, Res Inst 54, Shijiazhuang 050081, Peoples R China
基金
中国国家自然科学基金;
关键词
Dynamic scene analysis; sparse coding; topic model;
D O I
10.1109/TIP.2014.2363408
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Dynamic scene analysis has become a popular research area especially in video surveillance. The goal of this paper is to mine semantic motion patterns and detect abnormalities deviating from normal ones occurring in complex dynamic scenarios. To address this problem, we propose a data-driven and scene-independent approach, namely, Bilayer sparse topic model (BiSTM), where a given surveillance video is represented by a word-document hierarchical generative process. In this BiSTM, motion patterns are treated as latent topics sparsely distributed over low-level motion vectors, whereas a video clip can be sparsely reconstructed by a mixture of topics (motion pattern). In addition to capture the characteristic of extreme imbalance between numerous typical normal activities and few rare abnormalities in surveillance video data, a one-class constraint is directly imposed on the distribution of documents as a discriminant priori. By jointly learning topics and one-class document representation within a discriminative framework, the topic (pattern) space is more specific and explicit. An effective alternative iteration algorithm is presented for the model learning. Experimental results and comparisons on various public data sets demonstrate the promise of the proposed approach.
引用
收藏
页码:5198 / 5208
页数:11
相关论文
共 30 条
[11]  
Kratz L, 2009, PROC CVPR IEEE, P1446, DOI 10.1109/CVPRW.2009.5206771
[12]   What's going on? Discovering Spatio-Temporal Dependencies in Dynamic Scenes [J].
Kuettel, Daniel ;
Breitenstein, Michael D. ;
Van Gool, Luc ;
Ferrari, Vittorio .
2010 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2010, :1951-1958
[13]  
Li J, 2008, LECT NOTES COMPUT SC, V5305, P383
[14]   Anomaly Detection in Crowded Scenes [J].
Mahadevan, Vijay ;
Li, Weixin ;
Bhalodia, Viral ;
Vasconcelos, Nuno .
2010 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2010, :1975-1981
[15]  
Mehran R, 2009, PROC CVPR IEEE, P935, DOI 10.1109/CVPRW.2009.5206641
[16]   Scene Understanding by Statistical Modeling of Motion Patterns [J].
Saleemi, Imran ;
Hartung, Lance ;
Shah, Mubarak .
2010 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2010, :2069-2076
[17]   Estimating the support of a high-dimensional distribution [J].
Schölkopf, B ;
Platt, JC ;
Shawe-Taylor, J ;
Smola, AJ ;
Williamson, RC .
NEURAL COMPUTATION, 2001, 13 (07) :1443-1471
[18]   Learning patterns of activity using real-time tracking [J].
Stauffer, C ;
Grimson, WEL .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2000, 22 (08) :747-757
[19]  
Turek MW, 2010, LECT NOTES COMPUT SC, V6312, P664, DOI 10.1007/978-3-642-15552-9_48
[20]  
Varadarajan J., 2010, P NIPS WORKSH PRACT