Laboratoire
de Modélisation et Calcul



BLISS Project






Laboratoire
des Images et des Signaux
DEMONSTRATION of BLIND SOURCE SEPARATION

The demonstration implements three methods of BLIND SOURCE SEPARATION
for linear mixtures and a method for postnonlinear mixtures
The first method is based on the Gaussian Mutual Information, which exploits the non stationarity and the correlation of the sources and is implementated via a time-frequency analysis and a joint diagonalisation algorithm.
Reference papers:
Exploiting Source Non Stationary and Coloration in Blind Source Separation
(D. T. Pham and J.-F. Cardoso) Blind Separation of Instantaneous Mixtures of Non Stationary Sources
Blind Separation of Instantaneous Mixture of Sources via the Gaussian Mutual Information Criterion
Joint Approximate Diagonalization of Positive Definite Matrices
The second method is based on the minimisation of the Marginal Mutual Information criterion and exploits the non Gaussianity of the sources. The criterion is expressed in terms of the entropies which are estimated through a kernel method.
Reference papers:
Fast algorithms for Mutual Information Based Independent Component Analysis
Blind Separation of Instantaneous Mixture of Sources via an Independent Component Analysis
Fast Algorithm for Estimating Mutual Information, Entropies and Score Functions
The third method is also based on the minimisation of the Marginal Mutual Information criterion, but exploits the non stationarity of the sources as well.
Reference paper:
Blind Separation of Non Stationary Non Gaussian Sources.
The ponstnonlinear algorithm is based on the minimisation of the Marginal Mutual Information criterion.
Reference papers:
Blind source separation in postnonlinear mixtures
No Java demonstration since your brower dosen't support java
To download the Matlab Demonstration Code (nicer interface) click here



 Jitesh Shah