Speech recognition based on student's t-distribution derived from total Bayesian framework

Shinji Watanabe, Atsushi Nakamura

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

We introduce a robust classification method based on the Bayesian predictive distribution (Bayesian Predictive Classification, referred to as BPC) for speech recognition. We and others have recently proposed a total Bayesian framework named Variational Bayesian Estimation and Clustering for speech recognition (VBEC). VBEC includes the practical computation of approximate posterior distributions that are essential for BPC, based on variational Bayes (VB). BPC using VB posterior distributions (VB-BPC) provides an analytical solution for the predictive distribution as the Student's t-distribution, which can mitigate the over-training effects by marginalizing the model parameters of an output distribution. We address the sparse data problem in speech recognition, and show experimentally that VB-BPC is robust against data sparseness.

Original languageEnglish
Pages (from-to)970-980
Number of pages11
JournalIEICE Transactions on Information and Systems
VolumeE89-D
Issue number3
DOIs
Publication statusPublished - 2006
Externally publishedYes

Fingerprint

Speech recognition
Students

Keywords

  • Bayesian prediction
  • Speech recognition
  • Student's t-distribution
  • Total Bayesian framework VBEC

ASJC Scopus subject areas

  • Information Systems
  • Computer Graphics and Computer-Aided Design
  • Software

Cite this

Speech recognition based on student's t-distribution derived from total Bayesian framework. / Watanabe, Shinji; Nakamura, Atsushi.

In: IEICE Transactions on Information and Systems, Vol. E89-D, No. 3, 2006, p. 970-980.

Research output: Contribution to journalArticle

@article{85da1fd886914583bcc5f5847ec83ba8,
title = "Speech recognition based on student's t-distribution derived from total Bayesian framework",
abstract = "We introduce a robust classification method based on the Bayesian predictive distribution (Bayesian Predictive Classification, referred to as BPC) for speech recognition. We and others have recently proposed a total Bayesian framework named Variational Bayesian Estimation and Clustering for speech recognition (VBEC). VBEC includes the practical computation of approximate posterior distributions that are essential for BPC, based on variational Bayes (VB). BPC using VB posterior distributions (VB-BPC) provides an analytical solution for the predictive distribution as the Student's t-distribution, which can mitigate the over-training effects by marginalizing the model parameters of an output distribution. We address the sparse data problem in speech recognition, and show experimentally that VB-BPC is robust against data sparseness.",
keywords = "Bayesian prediction, Speech recognition, Student's t-distribution, Total Bayesian framework VBEC",
author = "Shinji Watanabe and Atsushi Nakamura",
year = "2006",
doi = "10.1093/ietisy/e89-d.3.970",
language = "English",
volume = "E89-D",
pages = "970--980",
journal = "IEICE Transactions on Information and Systems",
issn = "0916-8532",
publisher = "Maruzen Co., Ltd/Maruzen Kabushikikaisha",
number = "3",

}

TY - JOUR

T1 - Speech recognition based on student's t-distribution derived from total Bayesian framework

AU - Watanabe, Shinji

AU - Nakamura, Atsushi

PY - 2006

Y1 - 2006

N2 - We introduce a robust classification method based on the Bayesian predictive distribution (Bayesian Predictive Classification, referred to as BPC) for speech recognition. We and others have recently proposed a total Bayesian framework named Variational Bayesian Estimation and Clustering for speech recognition (VBEC). VBEC includes the practical computation of approximate posterior distributions that are essential for BPC, based on variational Bayes (VB). BPC using VB posterior distributions (VB-BPC) provides an analytical solution for the predictive distribution as the Student's t-distribution, which can mitigate the over-training effects by marginalizing the model parameters of an output distribution. We address the sparse data problem in speech recognition, and show experimentally that VB-BPC is robust against data sparseness.

AB - We introduce a robust classification method based on the Bayesian predictive distribution (Bayesian Predictive Classification, referred to as BPC) for speech recognition. We and others have recently proposed a total Bayesian framework named Variational Bayesian Estimation and Clustering for speech recognition (VBEC). VBEC includes the practical computation of approximate posterior distributions that are essential for BPC, based on variational Bayes (VB). BPC using VB posterior distributions (VB-BPC) provides an analytical solution for the predictive distribution as the Student's t-distribution, which can mitigate the over-training effects by marginalizing the model parameters of an output distribution. We address the sparse data problem in speech recognition, and show experimentally that VB-BPC is robust against data sparseness.

KW - Bayesian prediction

KW - Speech recognition

KW - Student's t-distribution

KW - Total Bayesian framework VBEC

UR - http://www.scopus.com/inward/record.url?scp=33645785890&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33645785890&partnerID=8YFLogxK

U2 - 10.1093/ietisy/e89-d.3.970

DO - 10.1093/ietisy/e89-d.3.970

M3 - Article

AN - SCOPUS:33645785890

VL - E89-D

SP - 970

EP - 980

JO - IEICE Transactions on Information and Systems

JF - IEICE Transactions on Information and Systems

SN - 0916-8532

IS - 3

ER -