Nonparametric density estimation for human pose tracking

Thomas Brox, B. Rosenhahn, U. Kersting, D. Cremers
Pattern Recognition (Proc. DAGM), Springer, LNCS, Vol.4174: 546-555, Apr. 2006
Abstract: The present paper considers the supplement of prior knowledge about joint angle configurations in the scope of 3-D human pose tracking. Training samples obtained from an industrial marker based tracking system are used for a nonparametric Parzen density estimation in the 12-dimensional joint configuration space. These learned probability densities constrain the image-driven joint angle estimates by drawing solutions towards familiar configurations. This prevents the method from producing unrealistic pose estimates due to unreliable image cues. Experiments on sequences with a human leg model reveal a considerably increased robustness, particularly in the presence of disturbed images and occlusions.

Other associated files : brox_dagm06.pdf [1.6MB]  

Images and movies


BibTex reference

  author       = "T. Brox and B. Rosenhahn and U. Kersting and D. Cremers",
  title        = "Nonparametric density estimation for human pose tracking",
  booktitle    = "Pattern Recognition (Proc. DAGM)",
  series       = "Lecture Notes in Computer Science",
  volume       = "4174",
  pages        = "546-555",
  month        = "Apr.",
  year         = "2006",
  publisher    = "Springer",
  url          = "http://lmb.informatik.uni-freiburg.de/Publications/2006/Bro06f"

Other publications in the database