AUTOMATIC PHONETIC FEATURE LABELLING OF CONTINUOUS SPEECH

MA Huckvale; IS Howard; WJ Barry

AUTOMATIC PHONETIC FEATURE LABELLING OF CONTINUOUS SPEECH

MA Huckvale, IS Howard, WJ Barry

School of Engineering, Computing and Mathematics

Research output: Contribution to journal › Conference proceedings published in a journal › peer-review

Abstract

This paper updates our previous work on the automatic phonetic feature analysis of speech. Previously we have described how a bank of feature-detectors can be used as a front-end to traditional speech recognition pattern matching algorithms, with increased performance in speaker-independent isolated word recognition over purely acoustic front-ends. In this paper we extend the feature analysis to continuous speech: describing the labelling methodology and the additional classification performance of neural network classifiers over the Bayes normal classifier.

Original language	English
Pages (from-to)	2565-2568
Number of pages	0
Journal	1st European Conference on Speech Communication and Technology, EUROSPEECH 1989
Volume	0
Issue number	0
Publication status	Published - 1 Jan 1989

Cite this

@article{e738dcfc44cf493a963ccc0c7eb8924c,

title = "AUTOMATIC PHONETIC FEATURE LABELLING OF CONTINUOUS SPEECH",

abstract = "This paper updates our previous work on the automatic phonetic feature analysis of speech. Previously we have described how a bank of feature-detectors can be used as a front-end to traditional speech recognition pattern matching algorithms, with increased performance in speaker-independent isolated word recognition over purely acoustic front-ends. In this paper we extend the feature analysis to continuous speech: describing the labelling methodology and the additional classification performance of neural network classifiers over the Bayes normal classifier.",

author = "MA Huckvale and IS Howard and WJ Barry",

year = "1989",

month = jan,

day = "1",

language = "English",

volume = "0",

pages = "2565--2568",

journal = "1st European Conference on Speech Communication and Technology, EUROSPEECH 1989",

number = "0",

}

TY - JOUR

T1 - AUTOMATIC PHONETIC FEATURE LABELLING OF CONTINUOUS SPEECH

AU - Huckvale, MA

AU - Howard, IS

AU - Barry, WJ

PY - 1989/1/1

Y1 - 1989/1/1

N2 - This paper updates our previous work on the automatic phonetic feature analysis of speech. Previously we have described how a bank of feature-detectors can be used as a front-end to traditional speech recognition pattern matching algorithms, with increased performance in speaker-independent isolated word recognition over purely acoustic front-ends. In this paper we extend the feature analysis to continuous speech: describing the labelling methodology and the additional classification performance of neural network classifiers over the Bayes normal classifier.

AB - This paper updates our previous work on the automatic phonetic feature analysis of speech. Previously we have described how a bank of feature-detectors can be used as a front-end to traditional speech recognition pattern matching algorithms, with increased performance in speaker-independent isolated word recognition over purely acoustic front-ends. In this paper we extend the feature analysis to continuous speech: describing the labelling methodology and the additional classification performance of neural network classifiers over the Bayes normal classifier.

M3 - Conference proceedings published in a journal

VL - 0

SP - 2565

EP - 2568

JO - 1st European Conference on Speech Communication and Technology, EUROSPEECH 1989

JF - 1st European Conference on Speech Communication and Technology, EUROSPEECH 1989

IS - 0

ER -

AUTOMATIC PHONETIC FEATURE LABELLING OF CONTINUOUS SPEECH

Abstract

Fingerprint

Cite this