Learning Switching Linear Models of Human Motion
Abstract
The human figure exhibits complex and rich dynamic behavior that is both nonlinear and time-varying. Effective models of human dynamics can be learned from motion capture data using switching linear dynamic system (SLDS) models. We present results for human motion synthe(cid:173) sis, classification, and visual tracking using learned SLDS models. Since exact inference in SLDS is intractable, we present three approximate in(cid:173) ference algorithms and compare their performance. In particular, a new variational inference algorithm is obtained by casting the SLDS model as a Dynamic Bayesian Network. Classification experiments show the superiority of SLDS over conventional HMM's for our problem domain.
Cite
Text
Pavlovic et al. "Learning Switching Linear Models of Human Motion." Neural Information Processing Systems, 2000.Markdown
[Pavlovic et al. "Learning Switching Linear Models of Human Motion." Neural Information Processing Systems, 2000.](https://mlanthology.org/neurips/2000/pavlovic2000neurips-learning/)BibTeX
@inproceedings{pavlovic2000neurips-learning,
title = {{Learning Switching Linear Models of Human Motion}},
author = {Pavlovic, Vladimir and Rehg, James M. and MacCormick, John},
booktitle = {Neural Information Processing Systems},
year = {2000},
pages = {981-987},
url = {https://mlanthology.org/neurips/2000/pavlovic2000neurips-learning/}
}