(with Dirk Ormoneit, Hedvig Sidenbladh, and Trevor Hastie)
We present methods for learning and tracking human motion in video. We estimate a statistical model of typical activities from a large set of 3D periodic human motion data by segmenting these data automatically into a sequence of ``cycles''. Then the mean and the principal components of the cycles are computed using a new algorithm that accounts for missing information and enforces smooth transitions between cycles. The learned temporal model provides a prior probability distribution over human motions that can be used in a Bayesian framework for tracking human subjects in complex monocular video sequences and recovering their 3D motion.
For more information and results on 3D human tracking (click here).
Image sequences and learned models of filter responses are available for download (click here).
Ormoneit, D., Sidenbladh, S., Black, M. J., Hastie, T., Learning and tracking cyclic human motion, Advances in Neural Information Processing Systems 13, Leen, Todd K. and Dietterich, Thomas G. and Tresp, Volker, Eds., The MIT Press, pp. 894-900, 2001. (pdf), (ps.gz).
Sidenbladh, H. and Black, M. J., Learning image statistics for Bayesian tracking, Int. Conf. on Computer Vision, ICCV-2001, Vancouver, BC, Vol. II, pp. 709-716. (postscript, 2.8MB)(pdf, 0.38MB).
Sidenbladh, H., Black, M. J., and Fleet, D.J., Stochastic tracking of 3D human figures using 2D image motion, European Conference on Computer Vision, D. Vernon (Ed.), Springer Verlag, LNCS 1843, Dublin, Ireland, pp. 702-718 June 2000. (postscript and pdf)
Sidenbladh, H., De la Torre, F., Black, M. J., A framework for modeling the appearance of 3D articulated figures, Int. Conf. on Automatic Face and Gesture Recognition, Grenoble, France, April 2000. (postscript), (abstract)