Online Segmentation of Actions Using Hidden Markov Models and Conceptional Relations of Daily Actions

In this paper, we propose a robust online action recognition algorithm with a segmentation scheme that detects start and end points of action occurrences. Specifically, the alogorithm estimates reliably what kind of actions occurring at present time. The algorithm has following characteristics. (1) The algorithm incorporates human knowledge about relations between action names in order to toughen the recognition, thus it labels robustly multiple action names at the same time. (2) The algorithm uses time-series Action Probability that represents the likelihood of each action occurrence at every frame time. The Action Probability is obtained from time-series human motion using support vector machine. (3) The algorithm can detect robustly and immediately the segmental points using classification technique with hidden Markov models (HMIs) . The experimental results using real motion capture data show that our algorithm not only prevents the system from making unnecessary segments due to the error of time-series Action Probability but also decreases effectively the latency for detecting the segmental points.

[1]  D. Rubin,et al.  Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper , 1977 .

[2]  Mari Ostendorf,et al.  From HMM's to segment models: a unified view of stochastic modeling for speech recognition , 1996, IEEE Trans. Speech Audio Process..

[3]  Christoph Bregler,et al.  Learning and recognizing human dynamics in video sequences , 1997, Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[4]  Robert P. W. Duin,et al.  Outlier Detection Using Classifier Instability , 1998, SSPR/SPR.

[5]  David Barber,et al.  Bayesian Classification With Gaussian Processes , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[6]  Bernhard Schölkopf,et al.  Support Vector Method for Novelty Detection , 1999, NIPS.

[7]  P. Bartlett,et al.  Probabilities for SV Machines , 2000 .

[8]  Vladimir Pavlovic,et al.  Learning Switching Linear Models of Human Motion , 2000, NIPS.

[9]  Michael E. Tipping.  Sparse Bayesian Learning and the Relevance Vector Machine , 2001, J. Mach. Learn. Res..

[10]  Bernhard Schölkopf,et al.  Learning with kernels , 2001 .

[11]  Steven Lemm,et al.  A Dynamic HMM for On-line Segmentation of Sequential Data , 2001, NIPS.

[12]  Kazuhito Yokoi,et al.  Imitating human dance motions through motion structure analysis , 2002, IEEE/RSJ International Conference on Intelligent Robots and Systems.

[13]  Padhraic Smyth,et al.  Segmental semi-markov models and applications to sequence analysis , 2002 .

[14]  Masamichi Shimosaka,et al.  Human-like action recognition system using features extracted by human , 2002, IEEE/RSJ International Conference on Intelligent Robots and Systems.

[15]  B. Schölkopf,et al.  Efficient face detection by a cascaded support–vector machine expansion , 2004, Proceedings of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences.

[16]  Masamichi Shimosaka,et al.  Hierarchical recognition of daily human actions based on Continuous Hidden Markov Models , 2004, Sixth IEEE International Conference on Automatic Face and Gesture Recognition, 2004. Proceedings..

[17]  Masamichi Shimosaka,et al.  Recognition of Actions in Daily Life and its Performance Adjustment Based on Support Vector Learning , 2004, Int. J. Humanoid Robotics.

[18]  David J. C. MacKay,et al.  Information Theory, Inference, and Learning Algorithms , 2004, IEEE Transactions on Information Theory.

[19]  Keiichi Kemmotsu,et al.  Recognizing human behaviors with vision sensors in a network robot system , 2006, Proceedings 2006 IEEE International Conference on Robotics and Automation, 2006. ICRA 2006..