Objective Pathological Voice Quality Assessment Based on HOS Features

被引：5

作者：

Lee, Ji-Yeoun ^{[1
]}

Jeong, Sangbae ^{[1
]}

Choi, Hong-Shik ^{[2
]}

Hahn, Minsoo ^{[1
]}

机构：

[1] Informat & Commun Univ, Taejon 305732, South Korea

[2] Yonsei Univ, Coll Med, Yongdong Severance Hosp, Dept Otorhinolaryngol,Inst Logoped & Phoniatr, Seoul, South Korea

来源：

IEICE TRANSACTIONS ON INFORMATION AND SYSTEMS | 2008年 / E91D卷 / 12期

关键词：

pathological voice quality assessment; higher-order statistics; classification and regression tree; GRBAS;

D O I：

10.1093/ietisy/e91-d.12.2888

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

This work proposes new features to improve the pathological voice quality classification performance. They are the means. the variances, and the perturbations of the higher-order statistics (HOS) such as the skewness and the kurtosis. The HOS-based features show meaningful differences among normal. grade 1, grade 2, and grade 3 voices, classified in the GRBAS scale. The jitter. the shimmer, the harmonic-to-noise ratio (HNR). and the variance of the short-time energy are utilized as the conventional features. The performances are measured by the classification and regression tree (CART) method. Specifically, the CART-based method by utilizing both the conventional features and the HOS-based ones shows its effectiveness in the pathological voice quality Measurement. with the classification accuracy of 87.8%.

引用

页码：2888 / 2891

页数：4

共 7 条

[1] Disordered speech assessment using automatic methods based on quantitative measures [J].