Presentation 2011-06-23
Bayesian speech recognition based on model structure integration
Sayaka SHIOTA, Kei HASHIMOTO, Yoshihiko NANKAKU, Keiichi TOKUDA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This paper proposes an acoustic modeling technique using multiple model structures based on a Bayesian framework for speech recognition. The Bayesian approach is a statistical technique for estimating reliable predictive distributions by marginalizing model parameters, and its effectiveness in speech recognition was reported. However, although the basic idea of the Bayesian approach is to treat all parameters as random variables, only one model structure is still selected in the conventional method. To improve the model complexity, multiple model structures should be used. The proposed method focuses on model structures integration based on the Bayesian framework in acoustic modeling. Furthermore, this paper apply the deterministic annealing EM algorithm to the optimization for constructing the appropriate acoustic models.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) speech recognition / acoustic modeling / Bayesian approach / model structure integration / deterministic annealing
Paper # SP2011-32
Date of Issue

Conference Information
Committee SP
Conference Date 2011/6/16(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 Speech (SP)
Language ENG
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Bayesian speech recognition based on model structure integration
Sub Title (in English)
Keyword(1) speech recognition
Keyword(2) acoustic modeling
Keyword(3) Bayesian approach
Keyword(4) model structure integration
Keyword(5) deterministic annealing
1st Author's Name Sayaka SHIOTA
1st Author's Affiliation Department of Computer Science and Engineering, Nagoya Institute of Technology()
2nd Author's Name Kei HASHIMOTO
2nd Author's Affiliation Department of Computer Science and Engineering, Nagoya Institute of Technology
3rd Author's Name Yoshihiko NANKAKU
3rd Author's Affiliation Department of Computer Science and Engineering, Nagoya Institute of Technology
4th Author's Name Keiichi TOKUDA
4th Author's Affiliation Department of Computer Science and Engineering, Nagoya Institute of Technology
Date 2011-06-23
Paper # SP2011-32
Volume (vol) vol.111
Number (no) 97
Page pp.pp.-
#Pages 6
Date of Issue