IEICE Trans - Blind Source Separation of Convolutive Mixtures of Speech in Frequency Domain

Blind Source Separation of Convolutive Mixtures of Speech in Frequency Domain

Shoji MAKINO
Hiroshi SAWADA
Ryo MUKAI
Shoko ARAKI

Publication
IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences Vol.E88-A No.7 pp.1640-1655
Publication Date: 2005/07/01
Online ISSN:
DOI: 10.1093/ietfec/e88-a.7.1640
Print ISSN: 0916-8508
Type of Manuscript: Special Section INVITED PAPER (Special Section on Multi-channel Acoustic Signal Processing)
Category:
Keyword:
blind source separation, convolutive mixtures, independent component analysis, frequency-domain BSS, microphone array, adaptive beamformer,

Full Text: PDF(1.7MB)>>

Summary:
This paper overviews a total solution for frequency-domain blind source separation (BSS) of convolutive mixtures of audio signals, especially speech. Frequency-domain BSS performs independent component analysis (ICA) in each frequency bin, and this is more efficient than time-domain BSS. We describe a sophisticated total solution for frequency-domain BSS, including permutation, scaling, circularity, and complex activation function solutions. Experimental results of 2

2 (moving sources), (#sources

#microphones) in a room are promising.

open access publishing via