No Access

RECOGNITION OF ACTIONS IN DAILY LIFE AND ITS PERFORMANCE ADJUSTMENT BASED ON SUPPORT VECTOR LEARNING

Graduate School of Information Science and Technology, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113–8656, Japan

Search for more papers by this author

MASAMICHI SHIMOSAKA

Graduate School of Information Science and Technology, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113–8656, Japan

Search for more papers by this author

TATSUYA HARADA

Graduate School of Information Science and Technology, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113–8656, Japan

Search for more papers by this author

, and

TOMOMASA SATO

Graduate School of Information Science and Technology, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113–8656, Japan

Search for more papers by this author

https://doi.org/10.1142/S0219843604000368Cited by:6 (Source: Crossref)

Abstract

This paper presents a recognition method for human actions in daily life. The system deals with actions related to regular human activity such as walking or lying down. The main features of the proposed method are: (i) simultaneous recognition, (ii) expressing lack of clarity in human recognition, (iii) defining similarities between two motions by utilizing kernel functions derived from expressions of actions based on human knowledge, (iv) robust learning capability based on support vector machine. Comparison with neural networks optimized by a back propagation algorithm and decision trees generated by C4.5 proves that the accuracy of recognition in the proposed method is superior to others. Recognizing actions in daily life robustly is expected to ensure smooth communication between humans and robots and to enhance support functionality in intelligent systems.

Keywords:

References

T. Starner and A. Pentland, Visual recognition of American sign language using hidden Markov models, Int. Workshop Automatic Face and Gesture Recognition (FGR) (IEEE Press, Zürich, Switzerland, 1995) pp. 189–194. Google Scholar
A. Wilson and A. Bobick, IEEE Trans. Pattern Anal. Mach. Intell. 21(9), 884 (1999), DOI: 10.1109/34.790429. Crossref, Web of Science, Google Scholar
T. Mori, K. Tsujioka and T. Sato, Human-like action recognition system on whole body motion-captured file, Proc. 2001 IEEE/RSJ Int. Conf. Intelligent Robots and Systems (IEEE Press, Hawaii, USA, 2001) pp. 2066–2073. Google Scholar
C. Cédras and M. Shah, Image Vision Comput. 13(2), 129 (1995). Crossref, Web of Science, Google Scholar
BVH File Format, http://www.biovision.com/bvh.html . Google Scholar
B. Schölkopf and A. Smola , Learning with Kernels ( MIT Press , Cambridge, MA , 2002 ) . Google Scholar
D. Mackay , Lecture Notes for a Tutorial at NIPS ( 1997 ) . Google Scholar
M. Tipping, J. Mach. Learn. Res. 1, 211 (2001), DOI: 10.1162/15324430152748236. Web of Science, Google Scholar
C. Cortes and V. Vapnik, Mach. Learn. 20(3), 273 (1995). Web of Science, Google Scholar
V. Vapnik , The Nature of Statistical Learning Theory (Statistics for Engineering and Information Science) ( Springer Verlag , 1995 ) . Crossref, Google Scholar
G. Causwenberghs and T. Poggio, Advances in Neural Information Processing 13, eds. T. Leen, T. Diettrich and V. Tresp (MIT Press, Cambridge, MA, 2001) pp. 409–415. Google Scholar
O. Chapelle and V. Vapnik, Advances in Neural Information Processing Systems 12, eds. S. Solla, T. Leen and K. Müller (MIT Press, Cambridge, MA, 2000) pp. 230–236. Google Scholar
J. Platt, Advances in Large Margin Classifiers, eds. A. Smolaet al. (MIT Press, Cambridge, Massachusetts, 1999) pp. 61–74. Google Scholar
ICS Action Database, http://www.ics.t.u-tokyo.ac.jp/action/ . Google Scholar
J. Quinlan, Mach. Learn. 1(1), 81 (1986). Google Scholar
C4.5 Release 8, http://www.cse.unsw.edu.au/quinlan/c4.5r8.tar.gz . Google Scholar