Advertisement

Using Hierarchical Temporal Memory for Vision-Based Hand Shape Recognition under Large Variations in Hand’s Rotation

  • Tomasz Kapuscinski
Part of the Lecture Notes in Computer Science book series (LNCS, volume 6114)

Abstract

Hierarchical Temporal Memory (HTM), a new computational paradigm based on cortical theory, has been applied to vision-based hand shape recognition under large variations in hand’s rotation. HTM’s abilities to build invariant object representations and solve ambiguities have been explored and quite promising results have been achieved for the difficult recognition task. The four-component edge orientation histograms calculated from the Canny edge images, have been proposed as the output of the HTM sensors. The two-layer HTM, with 16x16 nodes in the first layer and 8x8 in the second one, has been experimentally selected as the structure giving the best results. The 8 hand shapes, generated for 360 different rotations, have been recognized with efficiency up to 92%.

Keywords

hand gestures recognition hierarchical temporal memory human-computer interaction 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Kraiss, K.F.: Advanced Man-Machine Interaction. Springer, Berlin (2006)CrossRefGoogle Scholar
  2. 2.
    Athitsos, V., Sclaroff, S.: Estimating 3D Hand Pose from a Cluttered Image. In: Proc. of IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2003), vol. 2, pp. 432–440 (2003)Google Scholar
  3. 3.
    Guan, H., Chang, J.S., Chen, L., Feris, R.S., Turk, M.: Multi-view appearance-based 3d hand pose estimation. In: Proc. of Conference on Computer Vision and Pattern Recognition Workshop, pp. 154–155. IEEE Computer Society, Washington (2006)Google Scholar
  4. 4.
    Tanimoto, T., Hoshino, K.: Real Time Posture Estimation of Human Hand for Robot Hand Interface. In: Second International Symposium on Universal Communication, pp. 303–308. IEEE Computer Society, Los Alamitos (2008)CrossRefGoogle Scholar
  5. 5.
    Lee, S.U., Cohen, I.: 3D Hand Reconstruction from a Monocular View. In: 17th International Conference on Pattern Recognition (ICPR 2004), vol. 3, pp. 310–313 (2004)Google Scholar
  6. 6.
    Ong, E., Bowden, R.: A boosted classifier tree for hand shape detection. In: Proc. of 6th International Conference on Automatic Face and Gesture Recognition, pp. 889–894 (2004)Google Scholar
  7. 7.
    Stenger, B., Thayananthan, A., Torr, P., Cipolla, R.: Hand pose estimation using hierarchical detection. LNCS, pp. 105–116. Springer, Heidelberg (2004)Google Scholar
  8. 8.
    Hawkins, J.: Learn like a human. IEEE Spectrum (2007)Google Scholar
  9. 9.
    Hawkins, J., Blakeslee, S.: On Intelligence. Times Books, New York (2004)Google Scholar
  10. 10.
    Hawkins, J., Dileep, G.: Hierarchical Temporal Memory, Concepts, Theory and Terminology. Numenta Inc. (2006), http://www.numenta.com/Numenta_HTM_Concepts.pdf
  11. 11.
    Kapuscinski, T., Wysocki, M.: Using Hierarchical Temporal Memory for Recognition of Signed Polish Words. In: Kurzynski, M., Wozniak, M. (eds.) Computer Recognition Systems: Advances in Intelligent and Soft Computing, pp. 355–362. Springer, Heidelberg (2009)Google Scholar
  12. 12.
    Szczepankowski, B.: Sign language in school. WSiP, Warsaw (1988) (in Polish)Google Scholar
  13. 13.
    Dileep, G., Bobby, J.: The HTM Learning Algorithms. Numenta Inc. (2007), http://www.numenta.com/for-developers/education/algorithms.php
  14. 14.
    Virtual Hand Studio. 3D Virtual Figure Drawing, Anatomy Softwares and Anatomical models. D-AnatomyStore, CloudStars, http://d-anatomystore.com/
  15. 15.
    Numenta Platform for Intelligent Computing (NuPIC). Numenta Inc., http://www.numenta.com

Copyright information

© Springer-Verlag Berlin Heidelberg 2010

Authors and Affiliations

  • Tomasz Kapuscinski
    • 1
  1. 1.Department of Computer and Control EngineeringRzeszow University of TechnologyRzeszowPoland

Personalised recommendations