Pathological Voice Detection Using Efficient Combination of Heterogeneous Features

Ji-Yeoun LEE; Sangbae JEONG; Minsoo HAHN

doi:10.1093/ietisy/e91-d.2.367

Pathological Voice Detection Using Efficient Combination of Heterogeneous Features

Ji-Yeoun LEE, Sangbae JEONG, Minsoo HAHN

Full Text Views

0

Cite this

Summary :

Combination of mutually complementary features is necessary to cope with various changes in pattern classification between normal and pathological voices. This paper proposes a method to improve pathological/normal voice classification performance by combining heterogeneous features. Different combinations of auditory-based and higher-order features are investigated. Their performances are measured by Gaussian mixture models (GMMs), linear discriminant analysis (LDA), and a classification and regression tree (CART) method. The proposed classification method by using the CART analysis is shown to be an effective method for pathological voice detection, with a 92.7% classification performance rate. This is a noticeable improvement of 54.32% compared to the MFCC-based GMM algorithm in terms of error reduction.

Publication: IEICE TRANSACTIONS on Information Vol.E91-D No.2 pp.367-370

Publication Date: 2008/02/01

Publicized

Online ISSN: 1745-1361

DOI: 10.1093/ietisy/e91-d.2.367

Type of Manuscript: LETTER

Category: Speech and Hearing

Cite this

Copy

Ji-Yeoun LEE, Sangbae JEONG, Minsoo HAHN, "Pathological Voice Detection Using Efficient Combination of Heterogeneous Features" in IEICE TRANSACTIONS on Information, vol. E91-D, no. 2, pp. 367-370, February 2008, doi: 10.1093/ietisy/e91-d.2.367.
Abstract: Combination of mutually complementary features is necessary to cope with various changes in pattern classification between normal and pathological voices. This paper proposes a method to improve pathological/normal voice classification performance by combining heterogeneous features. Different combinations of auditory-based and higher-order features are investigated. Their performances are measured by Gaussian mixture models (GMMs), linear discriminant analysis (LDA), and a classification and regression tree (CART) method. The proposed classification method by using the CART analysis is shown to be an effective method for pathological voice detection, with a 92.7% classification performance rate. This is a noticeable improvement of 54.32% compared to the MFCC-based GMM algorithm in terms of error reduction.
URL: https://global.ieice.org/en_transactions/information/10.1093/ietisy/e91-d.2.367/_p

Copy

@ARTICLE{e91-d_2_367,
author={Ji-Yeoun LEE, Sangbae JEONG, Minsoo HAHN, },
journal={IEICE TRANSACTIONS on Information},
title={Pathological Voice Detection Using Efficient Combination of Heterogeneous Features},
year={2008},
volume={E91-D},
number={2},
pages={367-370},
abstract={Combination of mutually complementary features is necessary to cope with various changes in pattern classification between normal and pathological voices. This paper proposes a method to improve pathological/normal voice classification performance by combining heterogeneous features. Different combinations of auditory-based and higher-order features are investigated. Their performances are measured by Gaussian mixture models (GMMs), linear discriminant analysis (LDA), and a classification and regression tree (CART) method. The proposed classification method by using the CART analysis is shown to be an effective method for pathological voice detection, with a 92.7% classification performance rate. This is a noticeable improvement of 54.32% compared to the MFCC-based GMM algorithm in terms of error reduction.},
keywords={},
doi={10.1093/ietisy/e91-d.2.367},
ISSN={1745-1361},
month={February},}

Copy

TY - JOUR
TI - Pathological Voice Detection Using Efficient Combination of Heterogeneous Features
T2 - IEICE TRANSACTIONS on Information
SP - 367
EP - 370
AU - Ji-Yeoun LEE
AU - Sangbae JEONG
AU - Minsoo HAHN
PY - 2008
DO - 10.1093/ietisy/e91-d.2.367
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E91-D
IS - 2
JA - IEICE TRANSACTIONS on Information
Y1 - February 2008
AB - Combination of mutually complementary features is necessary to cope with various changes in pattern classification between normal and pathological voices. This paper proposes a method to improve pathological/normal voice classification performance by combining heterogeneous features. Different combinations of auditory-based and higher-order features are investigated. Their performances are measured by Gaussian mixture models (GMMs), linear discriminant analysis (LDA), and a classification and regression tree (CART) method. The proposed classification method by using the CART analysis is shown to be an effective method for pathological voice detection, with a 92.7% classification performance rate. This is a noticeable improvement of 54.32% compared to the MFCC-based GMM algorithm in terms of error reduction.
ER -