DMCA
Learning and classification of complex dynamics (2000)
Cached
Download Links
| Venue: | IEEE Transactions on Pattern Analysis and Machine Intelligence |
| Citations: | 89 - 2 self |
BibTeX
@ARTICLE{North00learningand,
author = {Ben North and Andrew Blake and Michael Isard and Jens Rittscher},
title = {Learning and classification of complex dynamics},
journal = {IEEE Transactions on Pattern Analysis and Machine Intelligence},
year = {2000},
volume = {22},
pages = {2000}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
AbstractÐStandard, exact techniques based on likelihood maximization are available for learning Auto-Regressive Process models of dynamical processes. The uncertainty of observations obtained from real sensors means that dynamics can be observed only approximately. Learning can still be achieved via ªEM-KºÐExpectation-Maximization (EM) based on Kalman Filtering. This cannot handle more complex dynamics, however, involving multiple classes of motion. A problem arises also in the case of dynamical processes observed visually: background clutter arising for example, in camouflage, produces non-Gaussian observation noise. Even with a single dynamical class, non-Gaussian observations put the learning problem beyond the scope of EM-K. For those cases, we show here how ªEM-CºÐbased on the CONDENSATION algorithm which propagates random ªparticle-sets,º can solve the learning problem. Here, learning in clutter is studied experimentally using visual observations of a hand moving over a desktop. The resulting learned dynamical model is shown to have considerable predictive value: When used as a prior for estimation of motion, the burden of computation in visual observation is significantly reduced. Multiclass dynamics are studied via visually observed juggling; plausible dynamical models have been found to emerge from the learning process, and accurate classification of motion has resulted. In practice, EM-C learning is computationally burdensome and the paper concludes with some discussion of computational complexity. Index TermsÐComputer vision, learning dynamics, Auto-Regressive Process, Expectation Maximization. 1
Keyphrases
complex dynamic learning problem dynamical process visual observation multiclass dynamic non-gaussian observation em-c learning auto-regressive process condensation algorithm non-gaussian observation noise multiple class learned dynamical model em-k expectation-maximization single dynamical class learning process background clutter plausible dynamical model likelihood maximization random particle-sets considerable predictive value abstract standard computational complexity index term computer vision auto-regressive process model accurate classification expectation maximization real sensor exact technique kalman filtering
