Convolutive blind source separation by fourth-order statistics

Fuxiang Wang*, Jun Zhang

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In this paper, we present a new method using fourth-order statistics for convolutive blind source separation. After a whitening process of the observed data, the separation of convolutive mixtures are transformed to find a semi-unitary demixing matrix. Then the semi-unitary demixing matrix is determined by joint diagonalization for a set of fourth-order joint cumulant matrices. Simulation results of speech separation demonstrate the effectiveness of the new approach.

Original languageEnglish
Title of host publicationICICS 2009 - Conference Proceedings of the 7th International Conference on Information, Communications and Signal Processing
DOIs
Publication statusPublished - 2009
Externally publishedYes
Event7th International Conference on Information, Communications and Signal Processing, ICICS 2009 - Macau Fisherman's Wharf, Macao
Duration: 8 Dec 200910 Dec 2009

Publication series

NameICICS 2009 - Conference Proceedings of the 7th International Conference on Information, Communications and Signal Processing

Conference

Conference7th International Conference on Information, Communications and Signal Processing, ICICS 2009
Country/TerritoryMacao
CityMacau Fisherman's Wharf
Period8/12/0910/12/09

Keywords

  • Blind source separation
  • Fourth-order statistics
  • Joint diagonalization

Fingerprint

Dive into the research topics of 'Convolutive blind source separation by fourth-order statistics'. Together they form a unique fingerprint.

Cite this