Blind Source Separation in the Frequency Domain: A Novel Solution to the Amplitude and the Permutation Indeterminacies

This paper deals with the separation of convolutive mixtures of statistically independent signals (sources) in the frequency domain. The convolutive mixture is decomposed in several problems of separating instantaneous mixtures which are independently solved. In addition, we propose a method to remove the indeterminacies which occur when all the individual separating systems do not extract the sources in the same order and with the same amplitude.We will show that both the permutation and the amplitude indeterminacies can be solved using second-order statistics when the sources are temporally-white.

[1]  Shiro Ikeda,et al.  A Method of Blind Separation Based on Temporal Structure of Signals , 1998, ICONIP.

[2]  Andrzej Cichocki,et al.  Stability Analysis of Learning Algorithms for Blind Source Separation , 1997, Neural Networks.

[3]  Jean-Francois Cardoso,et al.  Blind signal separation: statistical principles , 1998, Proc. IEEE.

[4]  Paris Smaragdis,et al.  Blind separation of convolved mixtures in the frequency domain , 1998, Neurocomputing.

[5]  Véronique Capdevielle,et al.  Blind adaptive separation of wide-band sources , 1996, 1996 IEEE International Conference on Acoustics, Speech, and Signal Processing Conference Proceedings.

[6]  L. Castedo,et al.  A neural network approach to blind source separation , 1997, Neural Networks for Signal Processing VII. Proceedings of the 1997 IEEE Signal Processing Society Workshop.

[7]  Jean-Louis Lacoume,et al.  Blind separation of wide-band sources in the frequency domain , 1995, 1995 International Conference on Acoustics, Speech, and Signal Processing.