3D Hand Tracking in a Stochastic Approximation Setting

D. Chik, J. Trumpf, and N. N. Schraudolph. 3D Hand Tracking in a Stochastic Approximation Setting. In 2^nd Workshop on Human Motion: Understanding, Modeling, Capture and Animation, 11^th IEEE Intl. Conf. Computer Vision (ICCV), pp. 136–151, Springer Verlag, Berlin, Rio de Janeiro, Brazil, 2007.

Download

1.3MB	207.6kB	2.5MB

Abstract

This paper introduces a hand tracking system with a theoretical proof of convergence. The tracking system follows a model-based approach and uses image-based cues, namely silhouettes and colour constancy. We show that, with the exception of a small set of parameter configurations, the cost function of our tracker has a well-behaved unique minimum. The convergence proof for the tracker relies on the convergence theory in stochastic approximation. We demonstrate that our tracker meets the sufficient conditions for stochastic approximation to hold locally. Experimental results on synthetic images generated from real hand motions show the feasibility of this approach.

BibTeX Entry

@inproceedings{ChiTruSch07,
     author = {Desmond Chik and Jochen Trumpf and Nicol N. Schraudolph},
      title = {\href{http://nic.schraudolph.org/pubs/ChiTruSch07.pdf}{
               3D Hand Tracking in a Stochastic Approximation Setting}},
      pages = {136--151},
     editor = {Ahmed Elgammal and Bodo Rosenhahn and Reinhard Klette},
  booktitle = {2$^{nd}$ Workshop on Human Motion: Understanding,
               Modeling, Capture and Animation, 11$^{th}$ IEEE
               Intl.\ Conf.\ Computer Vision (ICCV)},
     series = {\href{http://www.springer.de/comp/lncs/}{
               Lecture Notes in Computer Science}},
     volume =  4814,
  publisher = {\href{http://www.springer.de/}{Springer Verlag}, Berlin},
    address = {Rio de Janeiro, Brazil},
       year =  2007,
   b2h_type = {Other},
  b2h_topic = {>Stochastic Meta-Descent, Computer Vision},
   abstract = {
    This paper introduces a hand tracking system with a theoretical proof
    of convergence. The tracking system follows a model-based approach and
    uses image-based cues, namely silhouettes and colour constancy. We show
    that, with the exception of a small set of parameter configurations,
    the cost function of our tracker has a well-behaved unique minimum.
    The convergence proof for the tracker relies on the convergence theory
    in stochastic approximation. We demonstrate that our tracker meets the
    sufficient conditions for stochastic approximation to hold locally.
    Experimental results on synthetic images generated from real hand 
    motions show the feasibility of this approach. 
}}