Presentation 2010-12-21
A Voice Conversion Method Using Individuality of Sound Source
Akihito UEMURA, Jun TOYAMA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this paper, we describe a novel source conversion method for voice conversion. Toda et al. proposed voice conversion based on maximum likelihood estimation of spectral parameter trajectory. This method uses only vocal tract related parameters and dose not use source features. We treat the residuals of linear predictive coding as source signal to represent a target speaker individuality. Objective experimental results indicate that the power spectrum of the speech converted by our method is similar to a target speaker's one.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) voice conversion / source features / linear predictive coding error
Paper # NLC2010-26,SP2010-99
Date of Issue

Conference Information
Committee NLC
Conference Date 2010/12/13(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) A Voice Conversion Method Using Individuality of Sound Source
Sub Title (in English)
Keyword(1) voice conversion
Keyword(2) source features
Keyword(3) linear predictive coding error
1st Author's Name Akihito UEMURA
1st Author's Affiliation Graduate School of Information Science and Technology, Hokkaido University()
2nd Author's Name Jun TOYAMA
2nd Author's Affiliation Graduate School of Information Science and Technology, Hokkaido University
Date 2010-12-21
Paper # NLC2010-26,SP2010-99
Volume (vol) vol.110
Number (no) 356
Page pp.pp.-
#Pages 5
Date of Issue