Paper No: 81
A variational Bayesian algorithm for BSS problem with hidden Gauss-Markov Models for the sources
Author(s): Nadia
Bali, Ali Mohammad-Djafari
Abstract
In this paper we propose a Variational Bayesian estimation approach for Blind
Sources Separation problem, as an alternative method to MCMC. The data are
$M$ images and the sources are $N$ images which are assumed independent and
piecewise homogeneous. To insure these properties, we propose a piecewise
Gauss-Markov for the sources with a hidden classification variable which
is modeled via a Potts-Markov field. A few simulation results are given to
illustrate the performances of the proposed method and some comparison with
other methods (MCMC and VBICA) used for BSS. |
