Presentation 2004/10/22
Estimation of articulation from speech sounds considering the Quantal theory via a physiological articulatory model
Akikazu NISHIKIDO, Jianwu DANG,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Inverse estimation of articulation from speech sounds faces the problem of one-to-many. The method of inverse estimation of articulation from speech sounds with a physiological articulatory model can constrain the one-to-many problem to some extent by combining the spatial, dynamical and physiological constraint, but the problem basically is not solved. However, the one-to-many problem can be further reduced by distinguishing the global optimum and local optimum in articulatory space. To do so, we clarify the distribution of articulation in whole articulatory space and investigate the relation between the articulatory and acoustical characteristics of speech sounds. A reasonable result was obtained for the distribution of articulatory parameter in articulatory space, while the acoustical space is not completely consistent with the observations. Moreover, the relation between the distribution of articulatory point for Japanese vowel /a/ and the distribution of acoustical parameter indicated to support Quantal theory.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Speech Production / Articulatory Model / Physiological Model / Inverse Estimation / Quantal Theory
Paper # SP2004-69,WIT2004-50
Date of Issue

Conference Information
Committee WIT
Conference Date 2004/10/22(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Vice Chair

Paper Information
Registration To Well-being Information Technology(WIT)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Estimation of articulation from speech sounds considering the Quantal theory via a physiological articulatory model
Sub Title (in English)
Keyword(1) Speech Production
Keyword(2) Articulatory Model
Keyword(3) Physiological Model
Keyword(4) Inverse Estimation
Keyword(5) Quantal Theory
1st Author's Name Akikazu NISHIKIDO
1st Author's Affiliation School of Information Science, Japan Advanced Institute of Science and Technology()
2nd Author's Name Jianwu DANG
2nd Author's Affiliation ATR Human Information Science Labs
Date 2004/10/22
Paper # SP2004-69,WIT2004-50
Volume (vol) vol.104
Number (no) 389
Page pp.pp.-
#Pages 6
Date of Issue