Elliptical basis function networks and radial basis function networks for speaker verification: A comparative study

Man Wai Mak; C. K. Li

Elliptical basis function networks and radial basis function networks for speaker verification: A comparative study

The Hong Kong Polytechnic University

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

20 Citations (Scopus)

Abstract

It is well known that radial basis function (RBF) networks require a large number of function centers if the data to be modeled contain clusters with complicated shape. This paper proposes to overcome this problem by incorporating full covariance matrices into the RBF structure and to use the expectation-maximization (EM) algorithm to estimate the network parameters. The resulting networks, referred to as the elliptical basis function (EBF) networks, are applied to text-independent speaker verification. Experimental evaluations based on 258 speakers of the TIMIT corpus show that smaller size EBF networks with basis function parameters determined by the EM algorithm outperform the large RBF networks trained by the conventional approach.

Original language	English
Title of host publication	Proceedings of the International Joint Conference on Neural Networks
Publisher	IEEE
Pages	3034-3039
Number of pages	6
Publication status	Published - 1 Dec 1999
Event	International Joint Conference on Neural Networks (IJCNN'99) - Washington, DC, United States Duration: 10 Jul 1999 → 16 Jul 1999

Conference

Conference	International Joint Conference on Neural Networks (IJCNN'99)
Country/Territory	United States
City	Washington, DC
Period	10/07/99 → 16/07/99

ASJC Scopus subject areas

Software
Artificial Intelligence

Cite this

@inproceedings{bb0c4c441d5b473995318491603ad22d,

title = "Elliptical basis function networks and radial basis function networks for speaker verification: A comparative study",

abstract = "It is well known that radial basis function (RBF) networks require a large number of function centers if the data to be modeled contain clusters with complicated shape. This paper proposes to overcome this problem by incorporating full covariance matrices into the RBF structure and to use the expectation-maximization (EM) algorithm to estimate the network parameters. The resulting networks, referred to as the elliptical basis function (EBF) networks, are applied to text-independent speaker verification. Experimental evaluations based on 258 speakers of the TIMIT corpus show that smaller size EBF networks with basis function parameters determined by the EM algorithm outperform the large RBF networks trained by the conventional approach.",

author = "Mak, {Man Wai} and Li, {C. K.}",

year = "1999",

month = dec,

day = "1",

language = "English",

pages = "3034--3039",

booktitle = "Proceedings of the International Joint Conference on Neural Networks",

publisher = "IEEE",

address = "United States",

note = "International Joint Conference on Neural Networks (IJCNN'99) ; Conference date: 10-07-1999 Through 16-07-1999",

}

Elliptical basis function networks and radial basis function networks for speaker verification: A comparative study. / Mak, Man Wai; Li, C. K.
Proceedings of the International Joint Conference on Neural Networks. IEEE, 1999. p. 3034-3039.

Research output: Chapter in book / Conference proceeding › Conference article published in proceeding or book › Academic research › peer-review

TY - GEN

T1 - Elliptical basis function networks and radial basis function networks for speaker verification: A comparative study

AU - Mak, Man Wai

AU - Li, C. K.

PY - 1999/12/1

Y1 - 1999/12/1

N2 - It is well known that radial basis function (RBF) networks require a large number of function centers if the data to be modeled contain clusters with complicated shape. This paper proposes to overcome this problem by incorporating full covariance matrices into the RBF structure and to use the expectation-maximization (EM) algorithm to estimate the network parameters. The resulting networks, referred to as the elliptical basis function (EBF) networks, are applied to text-independent speaker verification. Experimental evaluations based on 258 speakers of the TIMIT corpus show that smaller size EBF networks with basis function parameters determined by the EM algorithm outperform the large RBF networks trained by the conventional approach.

AB - It is well known that radial basis function (RBF) networks require a large number of function centers if the data to be modeled contain clusters with complicated shape. This paper proposes to overcome this problem by incorporating full covariance matrices into the RBF structure and to use the expectation-maximization (EM) algorithm to estimate the network parameters. The resulting networks, referred to as the elliptical basis function (EBF) networks, are applied to text-independent speaker verification. Experimental evaluations based on 258 speakers of the TIMIT corpus show that smaller size EBF networks with basis function parameters determined by the EM algorithm outperform the large RBF networks trained by the conventional approach.

UR - http://www.scopus.com/inward/record.url?scp=0033313409&partnerID=8YFLogxK

M3 - Conference article published in proceeding or book

SP - 3034

EP - 3039

BT - Proceedings of the International Joint Conference on Neural Networks

PB - IEEE

T2 - International Joint Conference on Neural Networks (IJCNN'99)

Y2 - 10 July 1999 through 16 July 1999

ER -

Elliptical basis function networks and radial basis function networks for speaker verification: A comparative study

Abstract

Conference

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this