CONVOLUTIVE BSS USING A PENALIZED MUTUAL INFORMATION CRITERION (TuePmOR4)
Author(s) :
 Mohammed El Rhabi (Décom-LAM laboratory, University of Reims Champagne Ardennes, France) Guillaume Gelle (Décom-LAM laboratory, University of Reims Champagne Ardennes, France) Hassan Fenniri (Décom-LAM laboratory, University of Reims Champagne Ardennes, France) Georges Delaunay (Décom-LAM laboratory, University of Reims Champagne Ardennes, France)
 Abstract : The Blind Separation problem of linear time dependent mixtures is addressed in this paper. We have developed a novel algorithm based on the minimization of the mutual information plus a penalized term which ensures an {\it a priori} normalization of the estimated sources (outputs). The criterion minimization is done using a well known gradient approach. Finally, some numerical results are presented to illustrate the performance of the penalized algorithm comparing to the Babaie-Zadeh approach.