Wavelet de-noising for highly noisy source separation

The aim of this paper is to demonstrate that wavelet denoising processing is extremely attractive for efficient source separation of strong noisy mixtures. Systematic numerical simulations using source separation algorithms after wavelet de-noising are used to provide quantitative evaluations of the efficiency of the method. The cases of correlated Gaussian and non-Gaussian noise are investigated, which open the way to various practical applications.

Published in:
Proceedings - IEEE International Symposium on Circuits and Systems, 1, null, I/201-I/204
Presented at:
Phoenix, AZ
Swiss Federal Inst. of Technology, Lausanne, Switzerland Cited By: 3; Export Date: 14 August 2006; Source: Scopus CODEN: PICSD Language of Original Document: English Correspondence Address: Paraschiv-Ionescu, A.; Swiss Federal Inst. of Technology Lausanne, Switzerland References: Akuzawa, T., New fast factorization method for multivariate optimization and its realization as ICA algorithm http://www.mns.brain.riken.go.jp/'akuzawa; Attias, H., Independent factor analysis (1999) Neural Computation, 11, pp. 803-851; Buckheit, J., Donoho, D.L., Wavelab and reproductible research (1994) Wavelets in Statistics, pp. 55-82, In A. Antoniadis and G. Oppenheim, editors; Cichocki, A., Douglas, S.C., Amari, S., Robust techniques for independent component analysis with noisy data (1998) Neurocomputing, 22, pp. 113-129; Donoho, D.L., Johnstone, I.M., Adapting to unknown smoothness via wavelet shrinkage (1995) J. Am. Statist. Ass., 90, pp. 1200-1244; Donoho, D.L., Yu, T.P.Y., Nonlinear wavelet transforms based on median interpolation http://www- stat.stanford.edu/~donoho/Reports; Hyvarinen, A., Karhunen, J., Oja, E., (2001) Independent Component Analysis, John Wiley & Sons; Mallat, S.G., A theory of multiresolution signal decomposition: The wavelet representation (1989) IEEE Trans. Pattn. Anal. Mach. Intell., 11, pp. 674-693. Sponsors: IEEE
Other identifiers:
View record in Web of Science
Scopus: 2-s2.0-0036280709

 Record created 2006-11-30, last modified 2018-03-17

Rate this document:

Rate this document:
(Not yet reviewed)