Distributed Markov Chain Monte Carlo Particle Filtering

Abstract

Particle filters are computationally intensive and thus efficient parallelism is crucial to effective implementations, especially visual tracking in video sequences. Two schemes for pipelining data under high performance computing environment, including an alternative MCMC resampling algorithm, are proposed so as to improve diversity of resampling and… (More)
DOI: 10.1109/CSA.2009.5404264

Topics

6 Figures and Tables

Slides referencing similar topics