Voice activity detection based on a family of parametric distributions

被引：28

作者：

Shin, Jong Won ^{[1
]}

Chang, Joon-Hyuk

Kim, Nam Soo

机构：

[1] Seoul Natl Univ, Sch Elect Engn, Seoul 151742, South Korea

[2] Seoul Natl Univ, INMC, Seoul 151742, South Korea

[3] Inha Univ, Dept Elect Engn, Inchon 402751, South Korea

来源：

PATTERN RECOGNITION LETTERS | 2007年 / 28卷 / 11期

关键词：

statistical modeling; generalized gammal; voice activity detection;

D O I：

10.1016/j.patrec.2006.11.015

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 [模式识别与智能系统]; 0812 [计算机科学与技术]; 0835 [软件工程]; 1405 [智能科学与技术];

摘要：

In this letter, generalized gamma distribution (GFD) is introduced as a new statistical model of spectral distribution to be applied to the likelihood ratio test performed in voice activity detection (VAD). A gradient-based on-line algorithm is proposed to estimate the parameters of GFD according to the maximum likelihood criterion. Experimental results show that the VAD algorithm implemented based on GFD outperformed those adopting other parametric distributions. (c) 2007 Elsevier B.V. All rights reserved.

引用

页码：1295 / 1299

页数：5

共 17 条

[1]

[Anonymous], 1999, 301708 ETSI EN

[2]

A robust voice activity detector for wireless communications using soft computing [J].

Beritelli, F ;

Casale, S ;

Cavallaro, A .

IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, 1998, 16 (09) :1818-1829

[3]

CHANG JH, 2003, P EUR GEN SWITZ, P1065

[4]

CHANG JH, 2001, SPEECH ENHANCEMENT N, P1231

[5]

CHO YD, 2001, P IEEE INT C AC SPEE, V2, P7

[6]

Speech probability distribution [J].

Gazor, S ;

Zhang, W .

IEEE SIGNAL PROCESSING LETTERS, 2003, 10 (07) :204-207

[7]

HAIGH JA, 1993, TENCON'93: 1993 IEEE REGION 10 CONFERENCE ON COMPUTER, COMMUNICATION, CONTROL AND POWER ENGINEERING, VOL 3, P321, DOI 10.1109/TENCON.1993.327987

[8]

HOYT JD, 1994, INT CONF ACOUST SPEE, P237

[9]

*ITU T, 1996, G729 ITU T REC

[10]

Junqua J. C., 1991, P EUR, P1371

← 1 2 →