Effects of Bayesian predictive classification using variational Bayesian posteriors for sparse training data in speech recognition

Shinji Watanabe, Atsushi Nakamura

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

We introduce a robust classification method using Bayesian predictive distribution (Bayesian predictive classification, referred to as BPC) into speech recognition. We and others have recently proposed a total Bayesian framework for speech recognition, Variational Bayesian Estimation and Clustering for speech recognition (VBEC). VBEC includes an analytical derivation of approximate posterior distributions that are essential for BPC, based on variational Bayes (VB). BPC using VB posterior distributions (VB-BPC) can mitigate the over-training effects by marginalizing output distribution. We address the sparse data problem in speech recognition, and show how VB-BPC is robust against die data sparseness, experimentally.

Original languageEnglish
Title of host publication9th European Conference on Speech Communication and Technology, Eurospeech Interspeech
Pages1105-1108
Number of pages4
Publication statusPublished - 2005
Externally publishedYes
Event9th European Conference on Speech Communication and Technology - Lisbon
Duration: 2005 Sep 42005 Sep 8

Other

Other9th European Conference on Speech Communication and Technology
CityLisbon
Period05/9/405/9/8

Fingerprint

Speech recognition

ASJC Scopus subject areas

  • Engineering(all)

Cite this

Watanabe, S., & Nakamura, A. (2005). Effects of Bayesian predictive classification using variational Bayesian posteriors for sparse training data in speech recognition. In 9th European Conference on Speech Communication and Technology, Eurospeech Interspeech (pp. 1105-1108)

Effects of Bayesian predictive classification using variational Bayesian posteriors for sparse training data in speech recognition. / Watanabe, Shinji; Nakamura, Atsushi.

9th European Conference on Speech Communication and Technology, Eurospeech Interspeech. 2005. p. 1105-1108.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Watanabe, S & Nakamura, A 2005, Effects of Bayesian predictive classification using variational Bayesian posteriors for sparse training data in speech recognition. in 9th European Conference on Speech Communication and Technology, Eurospeech Interspeech. pp. 1105-1108, 9th European Conference on Speech Communication and Technology, Lisbon, 05/9/4.
Watanabe S, Nakamura A. Effects of Bayesian predictive classification using variational Bayesian posteriors for sparse training data in speech recognition. In 9th European Conference on Speech Communication and Technology, Eurospeech Interspeech. 2005. p. 1105-1108
Watanabe, Shinji ; Nakamura, Atsushi. / Effects of Bayesian predictive classification using variational Bayesian posteriors for sparse training data in speech recognition. 9th European Conference on Speech Communication and Technology, Eurospeech Interspeech. 2005. pp. 1105-1108
@inproceedings{d8b48e408f38454c9cc2fc3ac43902e7,
title = "Effects of Bayesian predictive classification using variational Bayesian posteriors for sparse training data in speech recognition",
abstract = "We introduce a robust classification method using Bayesian predictive distribution (Bayesian predictive classification, referred to as BPC) into speech recognition. We and others have recently proposed a total Bayesian framework for speech recognition, Variational Bayesian Estimation and Clustering for speech recognition (VBEC). VBEC includes an analytical derivation of approximate posterior distributions that are essential for BPC, based on variational Bayes (VB). BPC using VB posterior distributions (VB-BPC) can mitigate the over-training effects by marginalizing output distribution. We address the sparse data problem in speech recognition, and show how VB-BPC is robust against die data sparseness, experimentally.",
author = "Shinji Watanabe and Atsushi Nakamura",
year = "2005",
language = "English",
pages = "1105--1108",
booktitle = "9th European Conference on Speech Communication and Technology, Eurospeech Interspeech",

}

TY - GEN

T1 - Effects of Bayesian predictive classification using variational Bayesian posteriors for sparse training data in speech recognition

AU - Watanabe, Shinji

AU - Nakamura, Atsushi

PY - 2005

Y1 - 2005

N2 - We introduce a robust classification method using Bayesian predictive distribution (Bayesian predictive classification, referred to as BPC) into speech recognition. We and others have recently proposed a total Bayesian framework for speech recognition, Variational Bayesian Estimation and Clustering for speech recognition (VBEC). VBEC includes an analytical derivation of approximate posterior distributions that are essential for BPC, based on variational Bayes (VB). BPC using VB posterior distributions (VB-BPC) can mitigate the over-training effects by marginalizing output distribution. We address the sparse data problem in speech recognition, and show how VB-BPC is robust against die data sparseness, experimentally.

AB - We introduce a robust classification method using Bayesian predictive distribution (Bayesian predictive classification, referred to as BPC) into speech recognition. We and others have recently proposed a total Bayesian framework for speech recognition, Variational Bayesian Estimation and Clustering for speech recognition (VBEC). VBEC includes an analytical derivation of approximate posterior distributions that are essential for BPC, based on variational Bayes (VB). BPC using VB posterior distributions (VB-BPC) can mitigate the over-training effects by marginalizing output distribution. We address the sparse data problem in speech recognition, and show how VB-BPC is robust against die data sparseness, experimentally.

UR - http://www.scopus.com/inward/record.url?scp=33745190749&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33745190749&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:33745190749

SP - 1105

EP - 1108

BT - 9th European Conference on Speech Communication and Technology, Eurospeech Interspeech

ER -