@inbook{b1f07af53996467a9b55f530bb959ee7,
title = "Estimating the number of sources for frequency-domain blind source separation",
abstract = "Blind source separation (BSS) for convolutive mixtures can be performed efficiently in the frequency domain, where independent component analysis (ICA) is applied separately in each frequency bin. To solve the permutation problem of frequency-domain BSS robustly, information regarding the number of sources is very important. This paper presents a method for estimating the number of sources from convolutive mixtures of sources. The new method estimates the power of each source or noise component by using ICA and a scaling technique to distinguish sources and noises. Also, a reverberant component can be identified by calculating the correlation of component envelopes. Experimental results for up to three sources show that the proposed method worked well in a reverberant condition whose reverberation time was 200 ms.",
author = "Hiroshi Sawada and Stefan Winter and Ryo Mukai and Shoko Araki and Shoji Makino",
year = "2004",
doi = "10.1007/978-3-540-30110-3_78",
language = "English",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "610--617",
editor = "Puntonet, {Carlos G.} and Alberto Prieto",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
}