Coupling Semi-supervised Learning and Example Selection for Online Object Tracking

详细信息    查看全文

• 作者：Min Yang (17)
  Yuwei Wu (17)
  Mingtao Pei (17)
  Bo Ma (17)
  Yunde Jia (17)
  
  17. Beijing Laboratory of Intelligent Information Technology ; School of Computer Science ; Beijing Institute of Technology ; Beijing ; 100081 ; China
  
• 刊名：Lecture Notes in Computer Science
• 出版年：2015
• 出版时间：2015
• 年：2015
• 卷：9006
• 期：1
• 页码：476-491
• 全文大小：2,353 KB
• 参考文献：1. Ross, D, Lim, J, Lin, R, Yang, MH (2008) Incremental learning for robust visual tracking. Int. J. Comput. Vis. 77: pp. 125-141 CrossRef
  2. Kwon, J., Lee, K.: Visual tracking decomposition. In: CVPR, pp. 1269鈥?276 (2010)
  3. Mei, X., Ling, H.: Robust visual tracking using \(\ell 1\) minimization. In: ICCV, pp. 1鈥? (2009)
  4. Zhong, W., Lu, H., Yang, M.H.: Robust object tracking via sparsity-based collaborative model. In: CVPR, pp. 1838鈥?845 (2012)
  5. Jia, X., Lu, H., Yang, M.H.: Visual tracking via adaptive structural local sparse appearance model. In: CVPR, pp. 1822鈥?829 (2012)
  6. Wang, N., Wang, J., Yeung, D.Y.: Online robust non-negative dictionary learning for visual tracking. In: ICCV, pp. 657鈥?64 (2013)
  7. Wu, Y, Ma, B, Yang, M, Zhang, J, Jia, Y (2014) Metric learning based structural appearance model for robust visual tracking. IEEE Trans. Circuits Syst. Video Technol. 24: pp. 865-877 CrossRef
  8. Wang, D., Lu, H., Yang, M.H.: Least soft-thresold squares tracking. In: CVPR, pp. 2371鈥?378 (2013)
  9. Hare, S., Saffari, A., Torr, P.H.: Struck: structured output tracking with kernels. In: ICCV, pp. 263鈥?70 (2011)
  10. Zhang, K, Zhang, L, Yang, M-H Real-time compressive tracking. In: Fitzgibbon, A, Lazebnik, S, Perona, P, Sato, Y, Schmid, C eds. (2012) Computer Vision 鈥?ECCV 2012. Springer, Heidelberg, pp. 864-877 CrossRef
  11. Li, X., Shen, C., Dick, A.R., van den Hengel, A.: Learning compact binary codes for visual tracking. In: CVPR, pp. 2419鈥?426 (2013)
  12. Yao, R., Shi, Q., Shen, C., Zhang, Y., van den Hengel, A.: Part-based visual tracking with online latent structural learning. In: CVPR, pp. 2363鈥?370 (2013)
  13. Bai, Q., Wu, Z., Sclaroff, S., Betke, M., Monnier, C.: Randomized ensemble tracking. In: ICCV, pp. 2040鈥?047 (2013)
  14. Grabner, H, Leistner, C, Bischof, H Semi-supervised on-line boosting for robust tracking. In: Forsyth, D, Torr, P, Zisserman, A eds. (2008) Computer Vision 鈥?ECCV 2008. Springer, Heidelberg, pp. 234-247 CrossRef
  15. Babenko, B, Yang, MH, Belongie, S (2011) Robust object tracking with online multiple instance learning. IEEE Trans. Pattern Anal. Mach. Intell. 33: pp. 1619-1632 CrossRef
  16. Saffari, A, Leistner, C, Godec, M, Bischof, H Robust multi-view boosting with priors. In: Saffari, A, Leistner, C, Godec, M, Bischof, H eds. (2010) Computer Vision 鈥?ECCV 2010. Springer, Heidelberg, pp. 776-789 CrossRef
  17. Bai, Y., Tang, M.: Robust tracking via weakly supervised ranking SVM. In: CVPR, pp. 1854鈥?861 (2012)
  18. Gao, J., Xing, J., Hu, W., Maybank, S.: Discriminant tracking using tensor representation with semi-supervised improvement. In: ICCV (2013)
  19. Kalal, Z., Matas, J., Mikolajczyk, K.: P-N learning: bootstrapping binary classifiers by structural constraints. In: CVPR, pp. 49鈥?6 (2010)
  20. Supancic III, J.S., Ramanan, D.: Self-paced learning for long-term tracking. In: CVPR, pp. 2379鈥?386 (2013)
  21. Belkin, M, Niyogi, P, Sindhwani, V (2006) Manifold regularization: a geometric framework for learning from labeled and unlabeled examples. J. Mach. Learn. Res. 7: pp. 2399-2434
  22. Cohn, DA, Ghahramani, Z, Jordan, MI (1996) Active learning with statistical models. J. Artif. Intell. Res. 4: pp. 129-145
  23. Atkinson, AC, Donev, AN (2002) Optimum Experimental Designs. Oxford University Press, New York
  24. Santner, J., Leistner, C., Saffari, A., Pock, T., Bischof, H.: PROST: parallel robust online simple tracking. In: CVPR, pp. 723鈥?30 (2010)
  25. Yu, K., Bi, J., Tresp, V.: Active learning via transductive experimental design. In: ICML, pp. 1081鈥?088 (2006)
  26. He, X., Min, W., Cai, D., Zhou, K.: Laplacian optimal design for image retrieval. In: ACM SIGIR, pp. 119鈥?26 (2007)
  27. He, X (2010) Laplacian regularized d-optimal design for active learning and its application to image retrieval. IEEE Trans. Image Process. 19: pp. 254-263 CrossRef
  28. Zelnik-Manor, L., Perona, P.: Self-tuning spectral clustering. In: NIPS, pp. 1601鈥?608 (2004)
  29. Isard, M, Blake, A (1998) Condensation - conditional density propagation for visual tracking. Int. J. Comput. Vis. 29: pp. 5-28 CrossRef
  30. Wu, Y., Lim, J., Yang, M.H.: Online object tracking: a benchmark. In: CVPR, pp. 2411鈥?418 (2013)
  31. Dalal, N., Triggs, B.: Histograms of oriented gradients for human detection. In: CVPR, pp. 886鈥?93 (2005)
  
• 作者单位：Computer Vision -- ACCV 2014
• 丛书名：978-3-319-16816-6
• 刊物类别：Computer Science
• 刊物主题：Artificial Intelligence and Robotics
  Computer Communication Networks
  Software Engineering
  Data Encryption
  Database Management
  Computation by Abstract Devices
  Algorithm Analysis and Problem Complexity
  
• 出版者：Springer Berlin / Heidelberg
• ISSN：1611-3349

文摘

Training example collection is of great importance for discriminative trackers. Most existing algorithms use a sampling-and-labeling strategy, and treat the training example collection as a task that is independent of classifier learning. However, the examples collected directly by sampling are not intended to be useful for classifier learning. Updating the classifier with these examples might introduce ambiguity to the tracker. In this paper, we introduce an active example selection stage between sampling and labeling, and propose a novel online object tracking algorithm which explicitly couples the objectives of semi-supervised learning and example selection. Our method uses Laplacian Regularized Least Squares (LapRLS) to learn a robust classifier that can sufficiently exploit unlabeled data and preserve the local geometrical structure of feature space. To ensure the high classification confidence of the classifier, we propose an active example selection approach to automatically select the most informative examples for LapRLS. Part of the selected examples that satisfy strict constraints are labeled to enhance the adaptivity of our tracker, which actually provides robust supervisory information to guide semi-supervised learning. With active example selection, we are able to avoid the ambiguity introduced by an independent example collection strategy, and to alleviate the drift problem caused by misaligned examples. Comparison with the state-of-the-art trackers on the comprehensive benchmark demonstrates that our tracking algorithm is more effective and accurate.