Markovian blind image separation

Authors:
Shahram Hosseini;Rima Guidara;Yannick Deville;Christian Jutten
Affiliations:
Laboratoire Astrophysique de Toulouse-Tarbes (LATT), Observatoire Midi-Pyrénées – Université Paul Sabatier, Toulouse, France;Laboratoire Astrophysique de Toulouse-Tarbes (LATT), Observatoire Midi-Pyrénées – Université Paul Sabatier, Toulouse, France;Laboratoire Astrophysique de Toulouse-Tarbes (LATT), Observatoire Midi-Pyrénées – Université Paul Sabatier, Toulouse, France;Laboratoire des Images et des Signaux (LIS), UMR CNRS-INPG-UJF, Grenoble, France
Venue:
ICA'06 Proceedings of the 6th international conference on Independent Component Analysis and Blind Signal Separation
Year:
2006

Citing 4
Cited 2

Fast algorithms for mutual information based independent component analysis

IEEE Transactions on Signal Processing - Part I
A blind source separation technique using second-order statistics

IEEE Transactions on Signal Processing
Markovian source separation

IEEE Transactions on Signal Processing
Blind separation of mixture of independent sources through aquasi-maximum likelihood approach

IEEE Transactions on Signal Processing

Maximum likelihood blind image separation using nonsymmetrical half-plane Markov random fields

IEEE Transactions on Image Processing
Blind separation of non-stationary images using Markov models

ICA'07 Proceedings of the 7th international conference on Independent component analysis and signal separation

Quantified Score

Hi-index	0.00

Visualization

Abstract

We recently proposed a markovian image separation method. The proposed algorithm is however very time consuming so that it cannot be applied to large-size real-world images. In this paper, we propose two major modifications i.e. utilization of a low-cost parametric score function estimator and derivation of a modified equivariant version of Newton-Raphson algorithm for solving the estimating equations. These modifications make the algorithm much faster and allow us to perform more experiments with artificial and real data which are presented in the paper.