Share Email Print

Proceedings Paper

Multiscale blind source separation
Author(s): Pavel Kisilev; Michael Zibulevsky; Yehoshua Y. Zeevi; Barak A. Pearlmutter
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

The concern of the blind source separation problem is to extract the underlying source signals from a set of their linear mixtures, where the mixing matrix is unknown. It was discovered recently, that use of sparsity of source representation in some signal dictionary dramatically improves the quality of separation. In this work we use the property of multiscale transforms, such as wavelet or wavelet packets, to decompose signals into sets of local features with various degrees of sparsity. We use this intrinsic property for selecting the best (most sparse) subsets of features for further separation. Experiments with simulated signals, musical sounds and images demonstrate significant improvement of separation quality.

Paper Details

Date Published: 26 March 2001
PDF: 9 pages
Proc. SPIE 4391, Wavelet Applications VIII, (26 March 2001); doi: 10.1117/12.421206
Show Author Affiliations
Pavel Kisilev, Technion--Israel Institute of Technology (Israel)
Michael Zibulevsky, Technion--Israel Institute of Technology (Israel)
Yehoshua Y. Zeevi, Technion--Israel Institute of Technology (Israel)
Barak A. Pearlmutter, Univ. of New Mexico (United States)

Published in SPIE Proceedings Vol. 4391:
Wavelet Applications VIII
Harold H. Szu; David L. Donoho; Adolf W. Lohmann; William J. Campbell; James R. Buss, Editor(s)

© SPIE. Terms of Use
Back to Top