A SEGMENTAL APPROACH TO TEXT-INDEPENDENT SPEAKER VERIFICATION

J. Černocký; D. Petrovska-Delacrétaz; S. Pigeon; P. Verlinde; G. Chollet

A SEGMENTAL APPROACH TO TEXT-INDEPENDENT SPEAKER VERIFICATION

J. Černocký, D. Petrovska-Delacrétaz, S. Pigeon, P. Verlinde, G. Chollet

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

Abstract

Current text-independent speaker verification systems are usually based on modeling globally the probability density function (PDF) of the speaker feature vectors. In this paper, segmental approaches to text-independent speaker verification are discussed. Unlike the schemes based on Large Vocabulary Continuous Speech Recognition (LVCSR) with previously trained phone models, our systems are based on units derived in unsupervised manner using the ALISP (Automatic Language Independent Processing) tools. Speaker modeling is then done independently for each class of speech sounds. Among the techniques to merge the class-dependent scores, linear combination was tested and logistic regression and a method based on the Mixture of Experts technique are under investigation. The experimental results were obtained on the data from the NIST-NSA'98 campaign.

Original language	English
Pages	2207-2210
Number of pages	4
Publication status	Published - 1 Jan 1999
Externally published	Yes
Event	6th European Conference on Speech Communication and Technology, EUROSPEECH 1999 - Budapest, Hungary Duration: 5 Sept 1999 → 9 Sept 1999

Conference

Conference	6th European Conference on Speech Communication and Technology, EUROSPEECH 1999
Country/Territory	Hungary
City	Budapest
Period	5/09/99 → 9/09/99

Keywords

data fusion
segmental approach
text-independent speaker verification

Cite this

@conference{487a8aa1637a480196638033e7c8895b,

title = "A SEGMENTAL APPROACH TO TEXT-INDEPENDENT SPEAKER VERIFICATION",

abstract = "Current text-independent speaker verification systems are usually based on modeling globally the probability density function (PDF) of the speaker feature vectors. In this paper, segmental approaches to text-independent speaker verification are discussed. Unlike the schemes based on Large Vocabulary Continuous Speech Recognition (LVCSR) with previously trained phone models, our systems are based on units derived in unsupervised manner using the ALISP (Automatic Language Independent Processing) tools. Speaker modeling is then done independently for each class of speech sounds. Among the techniques to merge the class-dependent scores, linear combination was tested and logistic regression and a method based on the Mixture of Experts technique are under investigation. The experimental results were obtained on the data from the NIST-NSA'98 campaign.",

keywords = "data fusion, segmental approach, text-independent speaker verification",

author = "J. {\v C}ernock{\'y} and D. Petrovska-Delacr{\'e}taz and S. Pigeon and P. Verlinde and G. Chollet",

note = "Publisher Copyright: {\textcopyright} 1999 6th European Conference on Speech Communication and Technology, EUROSPEECH 1999. All rights reserved.; 6th European Conference on Speech Communication and Technology, EUROSPEECH 1999 ; Conference date: 05-09-1999 Through 09-09-1999",

year = "1999",

month = jan,

day = "1",

language = "English",

pages = "2207--2210",

}

TY - CONF

T1 - A SEGMENTAL APPROACH TO TEXT-INDEPENDENT SPEAKER VERIFICATION

AU - Černocký, J.

AU - Petrovska-Delacrétaz, D.

AU - Pigeon, S.

AU - Verlinde, P.

AU - Chollet, G.

PY - 1999/1/1

Y1 - 1999/1/1

N2 - Current text-independent speaker verification systems are usually based on modeling globally the probability density function (PDF) of the speaker feature vectors. In this paper, segmental approaches to text-independent speaker verification are discussed. Unlike the schemes based on Large Vocabulary Continuous Speech Recognition (LVCSR) with previously trained phone models, our systems are based on units derived in unsupervised manner using the ALISP (Automatic Language Independent Processing) tools. Speaker modeling is then done independently for each class of speech sounds. Among the techniques to merge the class-dependent scores, linear combination was tested and logistic regression and a method based on the Mixture of Experts technique are under investigation. The experimental results were obtained on the data from the NIST-NSA'98 campaign.

AB - Current text-independent speaker verification systems are usually based on modeling globally the probability density function (PDF) of the speaker feature vectors. In this paper, segmental approaches to text-independent speaker verification are discussed. Unlike the schemes based on Large Vocabulary Continuous Speech Recognition (LVCSR) with previously trained phone models, our systems are based on units derived in unsupervised manner using the ALISP (Automatic Language Independent Processing) tools. Speaker modeling is then done independently for each class of speech sounds. Among the techniques to merge the class-dependent scores, linear combination was tested and logistic regression and a method based on the Mixture of Experts technique are under investigation. The experimental results were obtained on the data from the NIST-NSA'98 campaign.

KW - data fusion

KW - segmental approach

KW - text-independent speaker verification

M3 - Paper

AN - SCOPUS:34147191768

SP - 2207

EP - 2210

T2 - 6th European Conference on Speech Communication and Technology, EUROSPEECH 1999

Y2 - 5 September 1999 through 9 September 1999

ER -

A SEGMENTAL APPROACH TO TEXT-INDEPENDENT SPEAKER VERIFICATION

Abstract

Conference

Keywords

Fingerprint

Cite this