Multiple Instance Learning of Visual Event Models

Abstract:

In this paper we propose a powerful visual event pattern learning method to address the issue of high-level video understanding. We first model the deformable temporal structure of the action event in videos by a temporal composition of several primitive motions. Moreover, we describe each action class by multiple temporal models to deal with the significant intra-class variability. We implement a multiple instance learning method to train the models in the weakly supervised setting. We have conducted experiments on three major benchmarks. The results are comparative to the state-of-the-arts.

• Periodical: Applied Mechanics and Materials (Volumes 321-324)
• Pages: 1030-1034
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.321-324.1030
• Online since: June 2013
• Authors: Jian Zhai Wu, De Wen Hu
• Keywords: Action Recognition, Multi-Instance Learning (MIL), Temporal Model, Video Comprehension

References

[1] P. Felzenszwalb, R. Girshick, D. McAllester and D. Ramanan: IEEE Trans Pattern Anal. Mach. Intell. Vol. 32 (2010), p.1627

DOI: 10.1109/tpami.2009.167

[2] P. Viola, J.C. Platt and C. Zhang: NIPS (2006)

[3] J. Friedman: Annals of Statistics Vol. 29 (2001), p.1189

[4] J.C. Niebles, C.W. Chen and L. Fei-Fei: ECCV (2010)

[5] M. Marszalek, I. Laptev and C. Schmid: CVPR (2009)

[6] J. Liu, J. Luo and M. Shah: CVPR (2009)

[7] Z. Lin, G. Hua and L.S. Davis: CVPR (2009)

[8] H. Wang, A. Klaser, C. Schimid and C. Liu: CVPR (2011)

[9] A. Vedaldi and A. Zisserman: CVPR (2010)

[10] W． Brendel and S. Todorovic: ECCV (2011)

[11] Y.G. Jiang, Q. Dai, X. Xue, W. Liu and C.W. Ngo: ECCV (2012)

[12] Q.V. Le, W.Y. Zou, S.Y. Yeung, A.Y. Ng: CVPR (2011)

[13] A. Gilbert, J. Illingworth, R. Bowden: IEEE Trans Pattern Anal. Mach. Intell. Vol. 33 (2011), p.883

[14] J. Liu, J. Luo, M. Shah: CVPR (2009)