DISCRIMINATIVE FEATURE AND MODEL DESIGN FOR AUTOMATIC SPEECH RECOGNITION

Mazin Rahim; Yoshua Bengio; Yann LeCun

DISCRIMINATIVE FEATURE AND MODEL DESIGN FOR AUTOMATIC SPEECH RECOGNITION

Mazin Rahim, Yoshua Bengio, Yann LeCun

Computer Science

Research output: Contribution to conference › Paper › peer-review

Abstract

A system for discriminative feature and model design is presented for automatic speech recognition. Training based on minimum classification error using a single objective function is applied for designing a set of parallel networks performing feature transformation and a set of hidden Markov models performing speech recognition. This paper compares the use of linear and non-linear functional transformations when applied to conventional recognition features, such as spectrum or cepstrum. It also provides a framework for integrated feature and model training when using class-specific transformations. Experimental results on telephone-based connected digit recognition are presented.

Original language	English (US)
Pages	75-78
Number of pages	4
State	Published - 1997
Event	5th European Conference on Speech Communication and Technology, EUROSPEECH 1997 - Rhodes, Greece Duration: Sep 22 1997 → Sep 25 1997

Conference

Conference	5th European Conference on Speech Communication and Technology, EUROSPEECH 1997
Country/Territory	Greece
City	Rhodes
Period	9/22/97 → 9/25/97

ASJC Scopus subject areas

Computer Science Applications
Software
Linguistics and Language
Communication

Cite this

@conference{3d6a213dc90c42b0879801a3e07be9ae,

title = "DISCRIMINATIVE FEATURE AND MODEL DESIGN FOR AUTOMATIC SPEECH RECOGNITION",

abstract = "A system for discriminative feature and model design is presented for automatic speech recognition. Training based on minimum classification error using a single objective function is applied for designing a set of parallel networks performing feature transformation and a set of hidden Markov models performing speech recognition. This paper compares the use of linear and non-linear functional transformations when applied to conventional recognition features, such as spectrum or cepstrum. It also provides a framework for integrated feature and model training when using class-specific transformations. Experimental results on telephone-based connected digit recognition are presented.",

author = "Mazin Rahim and Yoshua Bengio and Yann LeCun",

note = "Publisher Copyright: {\textcopyright} 1997 5th European Conference on Speech Communication and Technology, EUROSPEECH 1997. All rights reserved.; 5th European Conference on Speech Communication and Technology, EUROSPEECH 1997 ; Conference date: 22-09-1997 Through 25-09-1997",

year = "1997",

language = "English (US)",

pages = "75--78",

}

TY - CONF

T1 - DISCRIMINATIVE FEATURE AND MODEL DESIGN FOR AUTOMATIC SPEECH RECOGNITION

AU - Rahim, Mazin

AU - Bengio, Yoshua

AU - LeCun, Yann

PY - 1997

Y1 - 1997

N2 - A system for discriminative feature and model design is presented for automatic speech recognition. Training based on minimum classification error using a single objective function is applied for designing a set of parallel networks performing feature transformation and a set of hidden Markov models performing speech recognition. This paper compares the use of linear and non-linear functional transformations when applied to conventional recognition features, such as spectrum or cepstrum. It also provides a framework for integrated feature and model training when using class-specific transformations. Experimental results on telephone-based connected digit recognition are presented.

AB - A system for discriminative feature and model design is presented for automatic speech recognition. Training based on minimum classification error using a single objective function is applied for designing a set of parallel networks performing feature transformation and a set of hidden Markov models performing speech recognition. This paper compares the use of linear and non-linear functional transformations when applied to conventional recognition features, such as spectrum or cepstrum. It also provides a framework for integrated feature and model training when using class-specific transformations. Experimental results on telephone-based connected digit recognition are presented.

UR - http://www.scopus.com/inward/record.url?scp=85135185331&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85135185331&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:85135185331

SP - 75

EP - 78

T2 - 5th European Conference on Speech Communication and Technology, EUROSPEECH 1997

Y2 - 22 September 1997 through 25 September 1997

ER -

DISCRIMINATIVE FEATURE AND MODEL DESIGN FOR AUTOMATIC SPEECH RECOGNITION

Abstract

Conference

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this