Parallel Subspace Sampling for Particle Filtering in Dynamic Bayesian Networks

Authors:
Eva Besada-Portas;Sergey M. Plis;Jesus M. Cruz;Terran Lane
Affiliations:
University of New Mexico, Albuquerque, USA 87131;University of New Mexico, Albuquerque, USA 87131;Universidad Complutense de Madrid, Madrid, Spain 28040;University of New Mexico, Albuquerque, USA 87131
Venue:
ECML PKDD '09 Proceedings of the European Conference on Machine Learning and Knowledge Discovery in Databases: Part I
Year:
2009

Citing 5
Cited 2

A Probabilistic Exclusion Principle for Tracking Multiple Objects

International Journal of Computer Vision
Partitioned Sampling, Articulated Objects, and Interface-Quality Hand Tracking

ECCV '00 Proceedings of the 6th European Conference on Computer Vision-Part II
Rao-Blackwellised Particle Filtering for Dynamic Bayesian Networks

UAI '00 Proceedings of the 16th Conference on Uncertainty in Artificial Intelligence
Reducing particle filtering complexity for 3D motion capture using dynamic Bayesian networks

AAAI'08 Proceedings of the 23rd national conference on Artificial intelligence - Volume 3
Factored particles for scalable monitoring

UAI'02 Proceedings of the Eighteenth conference on Uncertainty in artificial intelligence

Adaptive parallel/serial sampling mechanisms for particle filtering in dynamic Bayesian networks

ECML PKDD'10 Proceedings of the 2010 European conference on Machine learning and knowledge discovery in databases: Part I
Swapping-based partitioned sampling for better complex density estimation: application to articulated object tracking

SUM'11 Proceedings of the 5th international conference on Scalable uncertainty management

Quantified Score

Hi-index	0.02

Visualization

Abstract

Monitoring the variables of real world dynamic systems is a difficult task due to their inherent complexity and uncertainty. Particle Filters (PF) perform that task, yielding probability distribution over the unobserved variables. However, they suffer from the curse of dimensionality problem: the number of particles grows exponentially with the dimensionality of the hidden state space. The problem is aggravated when the initial distribution of the variables is not well known, as happens in global localization problems. We present a new parallel PF for systems whose variable dependencies can be factored into a Dynamic Bayesian Network. The new algorithms significantly reduce the number of particles, while independently exploring different subspaces of hidden variables to build particles consistent with past history and measurements. We demonstrate this new PF approach on some complex dynamical system estimation problems, showing that our method successfully localizes and tracks hidden states in cases where traditional PFs fail.