Speaker verification using vector quantization and Hidden Markov Model

Mohd Zaizu Ilyas, Salina Abdul Samad, Aini Hussain, Khairul Anuar Ishak

Research output: Chapter in Book/Report/Conference proceedingConference contribution

10 Citations (Scopus)

Abstract

This paper presents a speaker verification system using a combination of Vector Quantization (VQ) and Hidden Markov Model (HMM) to improve the HMM performance. A Malay spoken digit database which contains 100 speakers is used for the testing and validation modules. It is shown that, by using the proposed combination technique, a total success rate (TSR) of 99.97% is achieved and it is an improvement of 11.24% in performance compared to HMM. For speaker verification, true speaker rejection rate, impostor acceptance rate and equal error rate (EER) are also improved significantly compared to HMM.

Original languageEnglish
Title of host publication2007 5th Student Conference on Research and Development, SCORED
DOIs
Publication statusPublished - 2007
Event2007 5th Student Conference on Research and Development, SCORED - Selangor
Duration: 11 Dec 200712 Dec 2007

Other

Other2007 5th Student Conference on Research and Development, SCORED
CitySelangor
Period11/12/0712/12/07

Fingerprint

performance
acceptance
Quantization
Hidden Markov model
Data base
Module
Acceptance
Testing

Keywords

  • Hidden Markov model
  • Speaker recognition
  • Speaker verification
  • Vector quantization

ASJC Scopus subject areas

  • Education
  • Management Science and Operations Research

Cite this

Ilyas, M. Z., Abdul Samad, S., Hussain, A., & Ishak, K. A. (2007). Speaker verification using vector quantization and Hidden Markov Model. In 2007 5th Student Conference on Research and Development, SCORED [4451419] https://doi.org/10.1109/SCORED.2007.4451419

Speaker verification using vector quantization and Hidden Markov Model. / Ilyas, Mohd Zaizu; Abdul Samad, Salina; Hussain, Aini; Ishak, Khairul Anuar.

2007 5th Student Conference on Research and Development, SCORED. 2007. 4451419.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Ilyas, MZ, Abdul Samad, S, Hussain, A & Ishak, KA 2007, Speaker verification using vector quantization and Hidden Markov Model. in 2007 5th Student Conference on Research and Development, SCORED., 4451419, 2007 5th Student Conference on Research and Development, SCORED, Selangor, 11/12/07. https://doi.org/10.1109/SCORED.2007.4451419
Ilyas MZ, Abdul Samad S, Hussain A, Ishak KA. Speaker verification using vector quantization and Hidden Markov Model. In 2007 5th Student Conference on Research and Development, SCORED. 2007. 4451419 https://doi.org/10.1109/SCORED.2007.4451419
Ilyas, Mohd Zaizu ; Abdul Samad, Salina ; Hussain, Aini ; Ishak, Khairul Anuar. / Speaker verification using vector quantization and Hidden Markov Model. 2007 5th Student Conference on Research and Development, SCORED. 2007.
@inproceedings{d58afed9cfc14f3aaaea5963dd1aeb28,
title = "Speaker verification using vector quantization and Hidden Markov Model",
abstract = "This paper presents a speaker verification system using a combination of Vector Quantization (VQ) and Hidden Markov Model (HMM) to improve the HMM performance. A Malay spoken digit database which contains 100 speakers is used for the testing and validation modules. It is shown that, by using the proposed combination technique, a total success rate (TSR) of 99.97{\%} is achieved and it is an improvement of 11.24{\%} in performance compared to HMM. For speaker verification, true speaker rejection rate, impostor acceptance rate and equal error rate (EER) are also improved significantly compared to HMM.",
keywords = "Hidden Markov model, Speaker recognition, Speaker verification, Vector quantization",
author = "Ilyas, {Mohd Zaizu} and {Abdul Samad}, Salina and Aini Hussain and Ishak, {Khairul Anuar}",
year = "2007",
doi = "10.1109/SCORED.2007.4451419",
language = "English",
isbn = "1424414709",
booktitle = "2007 5th Student Conference on Research and Development, SCORED",

}

TY - GEN

T1 - Speaker verification using vector quantization and Hidden Markov Model

AU - Ilyas, Mohd Zaizu

AU - Abdul Samad, Salina

AU - Hussain, Aini

AU - Ishak, Khairul Anuar

PY - 2007

Y1 - 2007

N2 - This paper presents a speaker verification system using a combination of Vector Quantization (VQ) and Hidden Markov Model (HMM) to improve the HMM performance. A Malay spoken digit database which contains 100 speakers is used for the testing and validation modules. It is shown that, by using the proposed combination technique, a total success rate (TSR) of 99.97% is achieved and it is an improvement of 11.24% in performance compared to HMM. For speaker verification, true speaker rejection rate, impostor acceptance rate and equal error rate (EER) are also improved significantly compared to HMM.

AB - This paper presents a speaker verification system using a combination of Vector Quantization (VQ) and Hidden Markov Model (HMM) to improve the HMM performance. A Malay spoken digit database which contains 100 speakers is used for the testing and validation modules. It is shown that, by using the proposed combination technique, a total success rate (TSR) of 99.97% is achieved and it is an improvement of 11.24% in performance compared to HMM. For speaker verification, true speaker rejection rate, impostor acceptance rate and equal error rate (EER) are also improved significantly compared to HMM.

KW - Hidden Markov model

KW - Speaker recognition

KW - Speaker verification

KW - Vector quantization

UR - http://www.scopus.com/inward/record.url?scp=50449098325&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=50449098325&partnerID=8YFLogxK

U2 - 10.1109/SCORED.2007.4451419

DO - 10.1109/SCORED.2007.4451419

M3 - Conference contribution

AN - SCOPUS:50449098325

SN - 1424414709

SN - 9781424414703

BT - 2007 5th Student Conference on Research and Development, SCORED

ER -