Two-stage online inference model for traffic pattern analysis and anomaly detection

详细信息    查看全文

• 作者：Hawook Jeong (1)
  Youngjoon Yoo (1)
  Kwang Moo Yi (1)
  Jin Young Choi (1)
  
• 关键词：Motion pattern analysis ; Video surveillance ; Anomaly detection ; Probabilistic topic model
• 刊名：Machine Vision and Applications
• 出版年：2014
• 出版时间：August 2014
• 年：2014
• 卷：25
• 期：6
• 页码：1501-1517
• 全文大小：2,871 KB
• 参考文献：1. Basharat, A., Gritai, A., Shah, M.: Learning object motion patterns for anomaly detection and improved object detection. In: IEEE Conference on CVPR (2008)
  2. Benezeth, Y., Jodoin, P.M., Saligrama, V.: Abnormality detection using low-level co-occurring events. Pattern Recogn. Lett. 32(3), 423鈥?31 (2011) CrossRef
  3. Bishop, C.M.: Pattern Recognition and Machine Learning (Information Science and Statistics). Springer-Verlag New York Inc, Secaucus, NJ, USA (2006)
  4. Blei, D.M., Ng, A.Y., Jordan, M.I.: Latent dirichlet allocation. JML. Res. 3, 993鈥?022 (2003)
  5. Canini, K.R., Shi, L., Neuroscience, H.W., Griffiths, T.L.: Online inference of topics with latent dirichlet allocation. In. In AI-STATS (2009)
  6. Duda, R.O., Hart, P.E., Stork, D.G.: Pattern Classification, 2nd edn. Wiley (2000)
  7. Emonet, R., Varadarajan, J., Odobez, J.M.: Extracting and locating temporal motifs in video scenes using a hierarchical non parametric bayesian model. In: IEEE Conference on CVPR, pp. 3233鈥?240 (2011)
  8. Griffiths, T.L., Steyvers, M.: Finding scientific topics. PNAS 101(suppl. 1), 5228鈥?235 (2004) CrossRef
  9. Hoffman, M., Blei, D.M., Bach, F.: Online learning for latent dirichlet allocation. In: NIPS (2010)
  10. Hospedales, T.M., Gong, S., Xiang, T.: A markov clustering topic model for mining behaviour in video. In: ICCV, pp. 1165鈥?172. IEEE (2009)
  11. Hu, W., Xiao, X., Fu, Z., Xie, D., Tan, T., Maybank, S.: A system for learning statistical motion patterns. IEEE Trans. Pattern Anal. Mach. Intell. 28(9), 1450鈥?464 (2006) CrossRef
  12. Kratz, L., Nishino, K.: Anomaly detection in extremely crowded scenes using spatio-temporal motion pattern models. 2013 IEEE Conference on Computer Vision and Pattern Recognition 0, pp. 1446鈥?453 (2009)
  13. Kuettel, D., Breitenstein, M.D., Van Gool, L., Ferrari, V.: What鈥檚 going on? discovering spatio-temporal dependencies in dynamic scenes. In: CVPR, pp. 1951鈥?958 (2010). doi:10.1109/CVPR.2010.5539869
  14. Machy, C., Desurmont, X., Delaigle, J.F., Bastide, A.: Introduction of cctv at level crossings with automatic detection of potentially dangerous situations (2007)
  15. Mahadevan, V., Li, W., Bhalodia, V., Vasconcelos, N.: Anomaly detection in crowded scenes. In: IEEE Conference on CVPR, pp. 1975鈥?981 (2010)
  16. Morris, B., Trivedi, M.: A survey of vision-based trajectory learning and analysis for surveillance. IEEE Trans. Circuits Systems Video Technol. 18(8), 1114鈥?127 (2008) CrossRef
  17. Morris, B., Trivedi, M.M.: Learning trajectory patterns by clustering: Experimental studies and comparative evaluation. In: CVPR, pp. 312鈥?19 (2009)
  18. Piciarelli, C., Foresti, G.L.: Online trajectory clustering for anomalous events detection. Pattern Recogn. Lett, 27(15), 1835鈥?842 (2006)
  19. Qin, Z., Shelton, C.R.: Improving multi-target tracking via social grouping. In: IEEE Conference on Computer Vision and Pattern Recognition (2012)
  20. Rodriguez, M., Ali, S., Kanade, T.: Tracking in unstructured crowded scenes. In: ICCV, pp. 1389鈥?396. IEEE (2009)
  21. Saleemi, I., Hartung, L., Shah, M.: Scene understanding by statistical modeling of motion patterns. In: CVPR, pp. 2069鈥?076. IEEE (2010)
  22. Saleemi, I., Shafique, K., Shah, M.: Probabilistic modeling of scene dynamics for applications in visual surveillance. IEEE Trans. PAMI 31(8), 1472鈥?485 (2009) CrossRef
  23. Stauffer, C., Grimson, W.E.L.: Adaptive background mixture models for real-time tracking. In: CVPR, pp. 2246鈥?252 (1999)
  24. Tomasi, C., Kanade, T.: Detection and tracking of point features. Tech. rep, IJCV (1991)
  25. UCSD: Anomaly dataset. http://www.svcl.ucsd.edu/projects/anoma-ly/dataset.html (2010)
  26. UMN: Crowd dataset. http://www.cs.ucf.edu/ramin/
  27. Varadarajan, J., Emonet, R., Odobez, J.: Bridging the past, present and future: Modeling scene activities from event relationships and global rules. In: IEEE Conference on CVPR, pp. 2096鈥?103 (2012)
  28. Walk, S., Majer, N., Schindler, K., Schiele, B.: New features and insights for pedestrian detection. In: Conference on CVPR. IEEE, IEEE, San Francisco (2010)
  29. Wang, B., Ye, M., Li, X., Zhao, F., Ding, J.: Abnormal crowd behavior detection using high-frequency and spatio-temporal features. Mach. Vis. Appl. 23(3), 501鈥?11 (2012) CrossRef
  30. Wang, X., Ma, K.T., Ng, G.W., Grimson, W.E.: Trajectory analysis and semantic region modeling using nonparametric hierarchical bayesian models. Int. J. Comput. Vis. 95(3), 287鈥?12 (2011) CrossRef
  31. Wang, X., Ma, X., Grimson, W.E.L.: Unsupervised activity perception in crowded and complicated scenes using hierarchical bayesian models. IEEE Trans. PAMI 31(3), 539鈥?55 (2009) CrossRef
  32. Wang, X., Tieu, K., Grimson, E.: Learning semantic scene models by trajectory analysis. In: Proceedings of the 9th ECCV - Volume Part III. ECCV鈥?6, pp. 110鈥?23. Springer, Berlin, Heidelberg (2006)
  33. Zhai, K., Boyd-Graber, J., Asadi, N., Alkhouja, M.: Mr. LDA: A flexible large scale topic modeling package using variational inference in mapreduce. In: ACM International Conference on World Wide Web (2012)
  34. Zhao, B., Fei-Fei, L., Xing, E.P.: Online detection of unusual events in videos via dynamic sparse coding. In: IEEE Conference on CVPR. Colorado Springs, CO (2011)
  
• 作者单位：Hawook Jeong (1)
  Youngjoon Yoo (1)
  Kwang Moo Yi (1)
  Jin Young Choi (1)
  
  1. Perception and Intelligence Laboratory, ASRI, Seoul National University, Room 413, Bldg 133, 599 Gwanak-ro, Gwanak-gu, Seoul, Korea
  
• ISSN：1432-1769

文摘

In this paper, we propose a method for modeling trajectory patterns with both regional and velocity observations through the probabilistic topic model. By embedding Gaussian models into the discrete topic model framework, our method uses continuous velocity as well as regional observations unlike existing approaches. In addition, the proposed framework combined with Hidden Markov Model can cover the temporal transition of the scene state, which is useful in checking a violation of the rule that some conflict topics (e.g. two cross-traffic patterns) should not occur at the same time. To achieve online learning even with the complexity of the proposed model, we suggest a novel learning scheme instead of collapsed Gibbs sampling. The proposed two-stage greedy learning scheme is not only efficient at reducing the search space but also accurate in a way that the accuracy of online learning becomes not worse than that of the batch learning. To validate the performance of our method, experiments were conducted on various datasets. Experimental results show that our model explains satisfactorily the trajectory patterns with respect to scene understanding, anomaly detection, and prediction.