Phonetically optimized speaker modeling for robust speaker recognition

Bong Jin Lee, Jeung Yoon Choi, Hong-Goo Kang

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

This paper proposes an efficient method to improve speaker recognition performance by dynamically controlling the ratio of phoneme class information. It utilizes the fact that each phoneme contains different amounts of speaker discriminative information that can be measured by mutual information. After classifying phonemes into five classes, the optimal ratio of each class in both training and testing processes is adjusted using a non-linear optimization technique, i.e., the Nelder-Mead method. Speaker identification results verify that the proposed method achieves 18% improvement in terms of error rate compared to a baseline system.

Original languageEnglish
JournalJournal of the Acoustical Society of America
Volume126
Issue number3
DOIs
Publication statusPublished - 2009 Sep 21

Fingerprint

phonemes
classifying
education
optimization
Phoneme
Modeling

All Science Journal Classification (ASJC) codes

  • Arts and Humanities (miscellaneous)
  • Acoustics and Ultrasonics

Cite this

@article{e8a031b536824e009f008db32f3c2bda,
title = "Phonetically optimized speaker modeling for robust speaker recognition",
abstract = "This paper proposes an efficient method to improve speaker recognition performance by dynamically controlling the ratio of phoneme class information. It utilizes the fact that each phoneme contains different amounts of speaker discriminative information that can be measured by mutual information. After classifying phonemes into five classes, the optimal ratio of each class in both training and testing processes is adjusted using a non-linear optimization technique, i.e., the Nelder-Mead method. Speaker identification results verify that the proposed method achieves 18{\%} improvement in terms of error rate compared to a baseline system.",
author = "Lee, {Bong Jin} and Choi, {Jeung Yoon} and Hong-Goo Kang",
year = "2009",
month = "9",
day = "21",
doi = "10.1121/1.3204765",
language = "English",
volume = "126",
journal = "Journal of the Acoustical Society of America",
issn = "0001-4966",
publisher = "Acoustical Society of America",
number = "3",

}

Phonetically optimized speaker modeling for robust speaker recognition. / Lee, Bong Jin; Choi, Jeung Yoon; Kang, Hong-Goo.

In: Journal of the Acoustical Society of America, Vol. 126, No. 3, 21.09.2009.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Phonetically optimized speaker modeling for robust speaker recognition

AU - Lee, Bong Jin

AU - Choi, Jeung Yoon

AU - Kang, Hong-Goo

PY - 2009/9/21

Y1 - 2009/9/21

N2 - This paper proposes an efficient method to improve speaker recognition performance by dynamically controlling the ratio of phoneme class information. It utilizes the fact that each phoneme contains different amounts of speaker discriminative information that can be measured by mutual information. After classifying phonemes into five classes, the optimal ratio of each class in both training and testing processes is adjusted using a non-linear optimization technique, i.e., the Nelder-Mead method. Speaker identification results verify that the proposed method achieves 18% improvement in terms of error rate compared to a baseline system.

AB - This paper proposes an efficient method to improve speaker recognition performance by dynamically controlling the ratio of phoneme class information. It utilizes the fact that each phoneme contains different amounts of speaker discriminative information that can be measured by mutual information. After classifying phonemes into five classes, the optimal ratio of each class in both training and testing processes is adjusted using a non-linear optimization technique, i.e., the Nelder-Mead method. Speaker identification results verify that the proposed method achieves 18% improvement in terms of error rate compared to a baseline system.

UR - http://www.scopus.com/inward/record.url?scp=70349095968&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=70349095968&partnerID=8YFLogxK

U2 - 10.1121/1.3204765

DO - 10.1121/1.3204765

M3 - Article

VL - 126

JO - Journal of the Acoustical Society of America

JF - Journal of the Acoustical Society of America

SN - 0001-4966

IS - 3

ER -