Presentation 2008-12-09
Speech Feature Extraction Using Constrained Nonnegative Matrix Factorization
Hyunsin PARK, Tetsuya TAKIGUCHI, Yasuo ARIKI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this paper, we propose a speech feature extraction approach using nonnegative matrix factorization (NMF), and a new initialization approach for NMF. NMF is excellent at extracting local feature using nonnegative constraint. This method has been applied to data of text, image, or sound. In this paper, we apply this method to speech feature extraction. Recently, initialization methods based on SVD or clustering methods for NMF have been proposed. In this paper, a new initialization method using correlation information for NMF is described. A sparseness of base vector is additionally constrained to NMF. In comparison experiment among initialization methods, our proposed method shows effectiveness at estimate error and word speech recognition. The new feature extraction method also shows effectiveness at word speech recognition compared to MFCC and features by principal component analysis or independent component analysis.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Nonnegative matrix factorization / Correlation propagation initialization / Speech feature extraction / Isolated word speech recognition
Paper # NLC2008-31,SP2008-86
Date of Issue

Conference Information
Committee NLC
Conference Date 2008/12/2(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 Natural Language Understanding and Models of Communication (NLC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Speech Feature Extraction Using Constrained Nonnegative Matrix Factorization
Sub Title (in English)
Keyword(1) Nonnegative matrix factorization
Keyword(2) Correlation propagation initialization
Keyword(3) Speech feature extraction
Keyword(4) Isolated word speech recognition
1st Author's Name Hyunsin PARK
1st Author's Affiliation Graduate School of Engineering, Kobe University()
2nd Author's Name Tetsuya TAKIGUCHI
2nd Author's Affiliation Graduate School of Engineering, Kobe University
3rd Author's Name Yasuo ARIKI
3rd Author's Affiliation Graduate School of Engineering, Kobe University
Date 2008-12-09
Paper # NLC2008-31,SP2008-86
Volume (vol) vol.108
Number (no) 337
Page pp.pp.-
#Pages 6
Date of Issue