Separation of speech signals for nonlinear mixtures
This paper shows an approach to recover original speech signals from their nonlinear mixtures. Using a geometric method that makes a piecewise linear approximation of the nonlinear mixing space, and the fact that the speech distributions are Laplacian or Gamma type, a set of slopes is obtained as a set of linear mixtures.
Unable to display preview. Download preview PDF.
- 1.A. Prieto, B. Prieto, C. G. Puntonet, P. M. Smith, A. Cañas, “Geometric separation of linear mixtures of sources: application to speech signals”. First International Congress on Independent Component Analisys and Signal Separation (ICA-99)”, 1999, 11–15 January, Aussois, France.Google Scholar
- 5.A. Taleb and C. Jutten. Nonlinear source separation: The Post-Nonlinear Mixtures. Proceedings of ESANN'97, 1997, pp. 279–284. D Facto Brussels.Google Scholar
- 6.P.A. Paratte and P. Robert, “Systèmes de mesure”, Traité d'Electricité, 1996, 17, Presses Polytechniques et Universitaires Romandes, Lausanne.Google Scholar