Speaker identification using ensembles of feature enhancement methods

Il Ho Yang; Min Seok Kim; Byung Min So; Myung Jae Kim; Ha Jin Yu

doi:10.1007/978-3-642-24082-9_74

Speaker identification using ensembles of feature enhancement methods

Il Ho Yang, Min Seok Kim, Byung Min So, Myung Jae Kim, Ha Jin Yu

Department of Computer Science and Engineering

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

Abstract

In this paper, we propose a classifier ensemble of various channel compensation and feature enhancement methods for robust speaker identification on various environments. The proposed ensemble system is constructed with 15 classifiers including three channel compensation methods (including CMS and variance normalization, and without compensation) and five feature enhancement methods (including PCA, kernel PCA, greedy kernel PCA, kernel multimodal discriminant analysis, and without enhancement). Experimental results show that the proposed ensemble system gives the highest average speaker identification rate in various environments (channels, noises, and sessions).

Original language	English
Title of host publication	Convergence and Hybrid Information Technology - 5th International Conference, ICHIT 2011, Proceedings
Pages	606-613
Number of pages	8
DOIs	https://doi.org/10.1007/978-3-642-24082-9_74
State	Published - 2011
Event	5th International Conference on Convergence and Hybrid Information Technology, ICHIT 2011 - Daejeon, Korea, Republic of Duration: 22 Sep 2011 → 24 Sep 2011

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	6935 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	5th International Conference on Convergence and Hybrid Information Technology, ICHIT 2011
Country/Territory	Korea, Republic of
City	Daejeon
Period	22/09/11 → 24/09/11

Keywords

classifier ensemble
greedy kernel PCA
kernel multimodal component analysis
speaker identification

Access to Document

10.1007/978-3-642-24082-9_74

Cite this

Yang, I. H., Kim, M. S., So, B. M., Kim, M. J., & Yu, H. J. (2011). Speaker identification using ensembles of feature enhancement methods. In Convergence and Hybrid Information Technology - 5th International Conference, ICHIT 2011, Proceedings (pp. 606-613). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6935 LNCS). https://doi.org/10.1007/978-3-642-24082-9_74

Yang, Il Ho ; Kim, Min Seok ; So, Byung Min et al. / Speaker identification using ensembles of feature enhancement methods. Convergence and Hybrid Information Technology - 5th International Conference, ICHIT 2011, Proceedings. 2011. pp. 606-613 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{912b435dff244b87b24bd85d6933ca80,

title = "Speaker identification using ensembles of feature enhancement methods",

abstract = "In this paper, we propose a classifier ensemble of various channel compensation and feature enhancement methods for robust speaker identification on various environments. The proposed ensemble system is constructed with 15 classifiers including three channel compensation methods (including CMS and variance normalization, and without compensation) and five feature enhancement methods (including PCA, kernel PCA, greedy kernel PCA, kernel multimodal discriminant analysis, and without enhancement). Experimental results show that the proposed ensemble system gives the highest average speaker identification rate in various environments (channels, noises, and sessions).",

keywords = "classifier ensemble, greedy kernel PCA, kernel multimodal component analysis, speaker identification",

author = "Yang, {Il Ho} and Kim, {Min Seok} and So, {Byung Min} and Kim, {Myung Jae} and Yu, {Ha Jin}",

year = "2011",

doi = "10.1007/978-3-642-24082-9_74",

language = "English",

isbn = "9783642240812",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

pages = "606--613",

booktitle = "Convergence and Hybrid Information Technology - 5th International Conference, ICHIT 2011, Proceedings",

note = "5th International Conference on Convergence and Hybrid Information Technology, ICHIT 2011 ; Conference date: 22-09-2011 Through 24-09-2011",

}

Yang, IH, Kim, MS, So, BM, Kim, MJ & Yu, HJ 2011, Speaker identification using ensembles of feature enhancement methods. in Convergence and Hybrid Information Technology - 5th International Conference, ICHIT 2011, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 6935 LNCS, pp. 606-613, 5th International Conference on Convergence and Hybrid Information Technology, ICHIT 2011, Daejeon, Korea, Republic of, 22/09/11. https://doi.org/10.1007/978-3-642-24082-9_74

Speaker identification using ensembles of feature enhancement methods. / Yang, Il Ho; Kim, Min Seok; So, Byung Min et al.
Convergence and Hybrid Information Technology - 5th International Conference, ICHIT 2011, Proceedings. 2011. p. 606-613 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 6935 LNCS).

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

TY - GEN

T1 - Speaker identification using ensembles of feature enhancement methods

AU - Yang, Il Ho

AU - Kim, Min Seok

AU - So, Byung Min

AU - Kim, Myung Jae

AU - Yu, Ha Jin

PY - 2011

Y1 - 2011

N2 - In this paper, we propose a classifier ensemble of various channel compensation and feature enhancement methods for robust speaker identification on various environments. The proposed ensemble system is constructed with 15 classifiers including three channel compensation methods (including CMS and variance normalization, and without compensation) and five feature enhancement methods (including PCA, kernel PCA, greedy kernel PCA, kernel multimodal discriminant analysis, and without enhancement). Experimental results show that the proposed ensemble system gives the highest average speaker identification rate in various environments (channels, noises, and sessions).

AB - In this paper, we propose a classifier ensemble of various channel compensation and feature enhancement methods for robust speaker identification on various environments. The proposed ensemble system is constructed with 15 classifiers including three channel compensation methods (including CMS and variance normalization, and without compensation) and five feature enhancement methods (including PCA, kernel PCA, greedy kernel PCA, kernel multimodal discriminant analysis, and without enhancement). Experimental results show that the proposed ensemble system gives the highest average speaker identification rate in various environments (channels, noises, and sessions).

KW - classifier ensemble

KW - greedy kernel PCA

KW - kernel multimodal component analysis

KW - speaker identification

UR - http://www.scopus.com/inward/record.url?scp=80054066185&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-24082-9_74

DO - 10.1007/978-3-642-24082-9_74

M3 - Conference contribution

AN - SCOPUS:80054066185

SN - 9783642240812

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 606

EP - 613

BT - Convergence and Hybrid Information Technology - 5th International Conference, ICHIT 2011, Proceedings

T2 - 5th International Conference on Convergence and Hybrid Information Technology, ICHIT 2011

Y2 - 22 September 2011 through 24 September 2011

ER -

Yang IH, Kim MS, So BM, Kim MJ, Yu HJ. Speaker identification using ensembles of feature enhancement methods. In Convergence and Hybrid Information Technology - 5th International Conference, ICHIT 2011, Proceedings. 2011. p. 606-613. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-642-24082-9_74

Speaker identification using ensembles of feature enhancement methods

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this