Combined speech recognition and speaker verification over the fixed and mobile telephone networks

Anastasis Kounoudes; Anixi Antonakoudi; Vassilis Kekatos; Philippos Peleties

Combined speech recognition and speaker verification over the fixed and mobile telephone networks

Anastasis Kounoudes, Anixi Antonakoudi, Vassilis Kekatos, Philippos Peleties

Electrical and Computer Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

3 Scopus citations

Abstract

A double-digit text-dependent speaker verification and text validation system is presented for use in telephone services. The system utilizes concatenated phoneme HMMs for both speech recognition and user authentication, and works in a soundprompted mode. Tests with Hidden Markov Models (HMMs) using Perceptual Linear Prediction (PLP) and Mel Frequency Cepstral Coefficients (MFCC) as well as Cepstral Mean Subtraction (CMS) are performed to assess their effect on recognition performance. The paper also studies the effects of various factors such as the length of the training data, the number of embedded re-estimations and Gaussian mixtures in training of the HMMs, the use of world models, bootstrapping, and user-depended thresholds on the performance of speech recognition and speaker verification.

Original language	English (US)
Title of host publication	Proceedings of the Third IASTED International Conference on Signal Processing, Pattern Recognition, and Applications
Pages	228-233
Number of pages	6
Volume	2006
State	Published - Dec 1 2006
Event	3rd IASTED International Conference on Signal Processing, Pattern Recognition, and Applications - Innsbruck, Austria Duration: Feb 15 2006 → Feb 17 2006

Other

Other	3rd IASTED International Conference on Signal Processing, Pattern Recognition, and Applications
Country/Territory	Austria
City	Innsbruck
Period	2/15/06 → 2/17/06

Keywords

Biometrics
Hidden markov models
Speaker verification
Text validation

Cite this

Combined speech recognition and speaker verification over the fixed and mobile telephone networks. / Kounoudes, Anastasis; Antonakoudi, Anixi; Kekatos, Vassilis; Peleties, Philippos.

Proceedings of the Third IASTED International Conference on Signal Processing, Pattern Recognition, and Applications. Vol. 2006 2006. p. 228-233.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Kounoudes, A, Antonakoudi, A, Kekatos, V & Peleties, P 2006, Combined speech recognition and speaker verification over the fixed and mobile telephone networks. in Proceedings of the Third IASTED International Conference on Signal Processing, Pattern Recognition, and Applications. vol. 2006, pp. 228-233, 3rd IASTED International Conference on Signal Processing, Pattern Recognition, and Applications, Innsbruck, Austria, 2/15/06.

@inproceedings{213cfcf3484b4229aba2dabead8096c4,

title = "Combined speech recognition and speaker verification over the fixed and mobile telephone networks",

abstract = "A double-digit text-dependent speaker verification and text validation system is presented for use in telephone services. The system utilizes concatenated phoneme HMMs for both speech recognition and user authentication, and works in a soundprompted mode. Tests with Hidden Markov Models (HMMs) using Perceptual Linear Prediction (PLP) and Mel Frequency Cepstral Coefficients (MFCC) as well as Cepstral Mean Subtraction (CMS) are performed to assess their effect on recognition performance. The paper also studies the effects of various factors such as the length of the training data, the number of embedded re-estimations and Gaussian mixtures in training of the HMMs, the use of world models, bootstrapping, and user-depended thresholds on the performance of speech recognition and speaker verification.",

keywords = "Biometrics, Hidden markov models, Speaker verification, Text validation",

author = "Anastasis Kounoudes and Anixi Antonakoudi and Vassilis Kekatos and Philippos Peleties",

year = "2006",

month = dec,

day = "1",

language = "English (US)",

isbn = "0889865477",

volume = "2006",

pages = "228--233",

booktitle = "Proceedings of the Third IASTED International Conference on Signal Processing, Pattern Recognition, and Applications",

note = "3rd IASTED International Conference on Signal Processing, Pattern Recognition, and Applications ; Conference date: 15-02-2006 Through 17-02-2006",

}

TY - GEN

T1 - Combined speech recognition and speaker verification over the fixed and mobile telephone networks

AU - Kounoudes, Anastasis

AU - Antonakoudi, Anixi

AU - Kekatos, Vassilis

AU - Peleties, Philippos

PY - 2006/12/1

Y1 - 2006/12/1

N2 - A double-digit text-dependent speaker verification and text validation system is presented for use in telephone services. The system utilizes concatenated phoneme HMMs for both speech recognition and user authentication, and works in a soundprompted mode. Tests with Hidden Markov Models (HMMs) using Perceptual Linear Prediction (PLP) and Mel Frequency Cepstral Coefficients (MFCC) as well as Cepstral Mean Subtraction (CMS) are performed to assess their effect on recognition performance. The paper also studies the effects of various factors such as the length of the training data, the number of embedded re-estimations and Gaussian mixtures in training of the HMMs, the use of world models, bootstrapping, and user-depended thresholds on the performance of speech recognition and speaker verification.

AB - A double-digit text-dependent speaker verification and text validation system is presented for use in telephone services. The system utilizes concatenated phoneme HMMs for both speech recognition and user authentication, and works in a soundprompted mode. Tests with Hidden Markov Models (HMMs) using Perceptual Linear Prediction (PLP) and Mel Frequency Cepstral Coefficients (MFCC) as well as Cepstral Mean Subtraction (CMS) are performed to assess their effect on recognition performance. The paper also studies the effects of various factors such as the length of the training data, the number of embedded re-estimations and Gaussian mixtures in training of the HMMs, the use of world models, bootstrapping, and user-depended thresholds on the performance of speech recognition and speaker verification.

KW - Biometrics

KW - Hidden markov models

KW - Speaker verification

KW - Text validation

UR - http://www.scopus.com/inward/record.url?scp=33847178385&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33847178385&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:33847178385

SN - 0889865477

SN - 9780889865471

VL - 2006

SP - 228

EP - 233

BT - Proceedings of the Third IASTED International Conference on Signal Processing, Pattern Recognition, and Applications

T2 - 3rd IASTED International Conference on Signal Processing, Pattern Recognition, and Applications

Y2 - 15 February 2006 through 17 February 2006

ER -

Combined speech recognition and speaker verification over the fixed and mobile telephone networks

Abstract

Other

Keywords

Other files and links

Fingerprint

Cite this