ePrints

Single-channel blind separation using L-1-sparse complex non-negative matrix factorization for acoustic signals

Lookup NU author(s): Phetcharat Parathai, Dr Wai Lok Woo, Emeritus Professor Satnam Dlay, Bin Gao

Downloads

Full text for this publication is not currently held within this repository. Alternative links are provided below where available.

Abstract

An innovative method of single-channel blind source separation is proposed. The proposed method is a complex-valued non-negative matrix factorization with probabilistically optimal L-1-norm sparsity. This preserves the phase information of the source signals and enforces the inherent structures of the temporal codes to be optimally sparse, thus resulting in more meaningful parts factorization. An efficient algorithm with closed-form expression to compute the parameters of the model including the sparsity has been developed. Real-time acoustic mixtures recorded from a single-channel are used to verify the effectiveness of the proposed method. (C) 2015 Acoustical Society of America

Publication metadata

Author(s): Parathai P, Woo WL, Dlay SS, Gao B

Publication type: Article

Publication status: Published

Journal: Journal of the Acoustical Society of America

Year: 2015

Volume: 137

Issue: 1

Pages: EL124-EL129

Print publication date: 01/01/2015

Online publication date: 06/01/2015

Acceptance date: 06/10/2014

ISSN (print): 0001-4966

ISSN (electronic): 1520-8524

Publisher: American Institute of Physics

URL: http://dx.doi.org/10.1121/1.4903913

DOI: 10.1121/1.4903913

Altmetrics

Share