Presentation 2013-10-11
Source number estimation under reverberant and underdetermined conditions based on clustering of source activity sequences
Nobutaka ITO, Ingrid JAFARI, Shoko ARAKI, Tomohiro NAKATANI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This paper proposes a novel method for source number estimation based on the clustering of source activity sequences. The method first derives source activity sequences through clustering-based blind source separation using more clusters than the maximum anticipated number of sources. It then classifies all source activity sequences simultaneously in all frequency bins, and finally obtains a source number estimate as the number of clusters in which activity concentrates. An experiment shows that the proposed method enumerates sources accurately even in reverberant and underdetermined conditions without using a sparse prior as per [1].
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Source number estimation / blind source separation / clustering / mixture model
Paper # EA2013-66
Date of Issue

Conference Information
Committee EA
Conference Date 2013/10/4(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Engineering Acoustics (EA)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Source number estimation under reverberant and underdetermined conditions based on clustering of source activity sequences
Sub Title (in English)
Keyword(1) Source number estimation
Keyword(2) blind source separation
Keyword(3) clustering
Keyword(4) mixture model
1st Author's Name Nobutaka ITO
1st Author's Affiliation NTT Communication Science Laboratories, NTT Corporatin()
2nd Author's Name Ingrid JAFARI
2nd Author's Affiliation NTT Communication Science Laboratories, NTT Corporatin
3rd Author's Name Shoko ARAKI
3rd Author's Affiliation NTT Communication Science Laboratories, NTT Corporatin
4th Author's Name Tomohiro NAKATANI
4th Author's Affiliation NTT Communication Science Laboratories, NTT Corporatin
Date 2013-10-11
Paper # EA2013-66
Volume (vol) vol.113
Number (no) 242
Page pp.pp.-
#Pages 5
Date of Issue