Multiple pronunciation dictionary using HMM-state confusion characteristics

Yumi Wakita, Harald Singer, Yoshinori Sagisaka

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

In this paper, we propose a POS (part-of-speech)-dependent multiple pronunciation dictionary generation method using HMM-state confusions spanning several phonemes. When used in a multi-pass search, a dictionary generated from the method makes it possible to recover missing words that are lost during the first pass of the search process in continuous speech recognition using a single pronunciation dictionary. The new pronunciations are added to a dictionary that considers the POS dependency of the confusion characteristics. Continuous word recognition experiments have confirmed that the best results are obtained when (1) confusions expressed by HMM-state sequences and (2) pronunciation variations considering the POS-dependent confusion characteristics are used.

Original languageEnglish
Pages (from-to)143-153
Number of pages11
JournalComputer Speech and Language
Volume13
Issue number2
DOIs
Publication statusPublished - 1999 Apr
Externally publishedYes

Fingerprint

Confusion
Glossaries
dictionary
Continuous speech recognition
Dependent
Speech Recognition
Dictionary
experiment
Experiment
Speech
Experiments

ASJC Scopus subject areas

  • Signal Processing
  • Electrical and Electronic Engineering
  • Experimental and Cognitive Psychology
  • Linguistics and Language

Cite this

Multiple pronunciation dictionary using HMM-state confusion characteristics. / Wakita, Yumi; Singer, Harald; Sagisaka, Yoshinori.

In: Computer Speech and Language, Vol. 13, No. 2, 04.1999, p. 143-153.

Research output: Contribution to journalArticle

@article{f4fd9e1669ef4899bb7217ef7918159d,
title = "Multiple pronunciation dictionary using HMM-state confusion characteristics",
abstract = "In this paper, we propose a POS (part-of-speech)-dependent multiple pronunciation dictionary generation method using HMM-state confusions spanning several phonemes. When used in a multi-pass search, a dictionary generated from the method makes it possible to recover missing words that are lost during the first pass of the search process in continuous speech recognition using a single pronunciation dictionary. The new pronunciations are added to a dictionary that considers the POS dependency of the confusion characteristics. Continuous word recognition experiments have confirmed that the best results are obtained when (1) confusions expressed by HMM-state sequences and (2) pronunciation variations considering the POS-dependent confusion characteristics are used.",
author = "Yumi Wakita and Harald Singer and Yoshinori Sagisaka",
year = "1999",
month = "4",
doi = "10.1006/csla.1998.0116",
language = "English",
volume = "13",
pages = "143--153",
journal = "Computer Speech and Language",
issn = "0885-2308",
publisher = "Academic Press Inc.",
number = "2",

}

TY - JOUR

T1 - Multiple pronunciation dictionary using HMM-state confusion characteristics

AU - Wakita, Yumi

AU - Singer, Harald

AU - Sagisaka, Yoshinori

PY - 1999/4

Y1 - 1999/4

N2 - In this paper, we propose a POS (part-of-speech)-dependent multiple pronunciation dictionary generation method using HMM-state confusions spanning several phonemes. When used in a multi-pass search, a dictionary generated from the method makes it possible to recover missing words that are lost during the first pass of the search process in continuous speech recognition using a single pronunciation dictionary. The new pronunciations are added to a dictionary that considers the POS dependency of the confusion characteristics. Continuous word recognition experiments have confirmed that the best results are obtained when (1) confusions expressed by HMM-state sequences and (2) pronunciation variations considering the POS-dependent confusion characteristics are used.

AB - In this paper, we propose a POS (part-of-speech)-dependent multiple pronunciation dictionary generation method using HMM-state confusions spanning several phonemes. When used in a multi-pass search, a dictionary generated from the method makes it possible to recover missing words that are lost during the first pass of the search process in continuous speech recognition using a single pronunciation dictionary. The new pronunciations are added to a dictionary that considers the POS dependency of the confusion characteristics. Continuous word recognition experiments have confirmed that the best results are obtained when (1) confusions expressed by HMM-state sequences and (2) pronunciation variations considering the POS-dependent confusion characteristics are used.

UR - http://www.scopus.com/inward/record.url?scp=0033106613&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0033106613&partnerID=8YFLogxK

U2 - 10.1006/csla.1998.0116

DO - 10.1006/csla.1998.0116

M3 - Article

AN - SCOPUS:0033106613

VL - 13

SP - 143

EP - 153

JO - Computer Speech and Language

JF - Computer Speech and Language

SN - 0885-2308

IS - 2

ER -