Advertisement

The Effect of Multiple Training Sequences on HMM Classification of Motion Capture Gesture Data

  • Michał Romaszewski
  • Przemysław Głomb
Part of the Advances in Intelligent and Soft Computing book series (AINSC, volume 95)

Abstract

Hidden Markov Models (HMM) have been successfully applied to classification of various types of data, including human gestures. However, finding an optimal size of the training set for HMMs trained with multiple sequences and choosing a set of parameters for high performance is not a trivial task. We would like to address those issues by presenting results obtained using classifier based on HMMand Vector Quantisation applied to the set of a human gesture recordings.We use HMM as a model of a single gesture, and assess its recognition performance for multiple data sequences consisting of repetitions of selected gestures, performed by different persons with varying speed of movement. Additionally we intend to verify a reference database of 22 gestures for use in future experiments.

Keywords

gesture recognition gesture interface Vector Quantisation HMM gesture database 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Axenbeck, T., Bennewitz, M., Behnke, S., Burgard, W.: Recognizing complex, parameterized gestures from monocular image sequences. In: 8th IEEE-RAS International Conference on Humanoid Robots, Humanoids 2008, pp. 687–692 (2008)Google Scholar
  2. 2.
    Cholewa, M., Głomb, P.: Gesture data modeling and classification based on critical points approximation. Accepted for 7th International Conference on Computer Recognition Systems CORES 2011 (2011)Google Scholar
  3. 3.
    Elmezain, M., Al Hamadi, A., Appenrodt, J., Michaelis, B.: A hidden markov model-based continuous gesture recognition system for hand motion trajectory. In: ICPR, pp. 1–4 (2008)Google Scholar
  4. 4.
    Głomb, P., Romaszewski, M., Opozda, S., Sochan, A.: Choosing and modeling gesture database for natural user interface. Accepted for 7th International Conference on Computer Recognition Systems CORES 2011 (2011)Google Scholar
  5. 5.
    Linde, Y., Buzo, A., Gray, R.: An algorithm for vector quantizer design. IEEE Transactions on Communications 28(1), 84–95 (1980)CrossRefGoogle Scholar
  6. 6.
    McNeill, D.: Hand and Mind: What Gestures Reveal about Thought. The University of Chicago Press, Chicago (1992)Google Scholar
  7. 7.
    Elisabetta Farella Piero Zappi, Bojan Milosevic and Luca Benini. Hidden markov model based gesture recognition on low-cost, low-power tangible user interfaces. Entertainment Computing, 1(2):75–78, 2009.CrossRefGoogle Scholar
  8. 8.
    Quek, F., McNeill, D., Bryll, R., Duncan, S., Ma, X.-F., Kirbas, C., McCullough, K.E., Ansari, R.: Multimodal human discourse: gesture and speech. ACM Trans. Comput.-Hum. Interact. 9, 171–193 (2002)CrossRefGoogle Scholar
  9. 9.
    Rabiner, L.R.: A tutorial on hidden markov models and selected applications in speech recognition. Proceedings of the IEEE, 257–286 (1989)Google Scholar
  10. 10.
    van Oosten, J.-P.: Can markov properties be learned by hidden markov modelling algorithms? Master’s thesis, University of Groningen, The Netherlands (2010)Google Scholar
  11. 11.
    Yang, H.-D., Park, A.-Y., Lee, S.-W.: Human-robot interaction by whole body gesture spotting and recognition. In: International Conference on Pattern Recognition, vol. 4, pp. 774–777 (2006)Google Scholar
  12. 12.
    Yang, J., Xu, Y.: Hidden markov model for gesture recognition. Technical Report CMU-RI-TR-94-10, Robotics Institute, Pittsburgh, PA (May 1994)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2011

Authors and Affiliations

  • Michał Romaszewski
    • 1
  • Przemysław Głomb
    • 1
  1. 1.Institute of Theoretical and Applied Informatics of PASGliwicePoland

Personalised recommendations