A Gaussian/Laplacian hybrid statistical voice activity detector for line spectral frequency-based speech coders

Author

Othman, H. ; Aboulnasr, T.

Author_Institution

Sch. of Inf. Technol. & Eng., Ottawa Univ., Ont., Canada

Volume

2

fYear

2003

fDate

27-30 Dec. 2003

Firstpage

693

Abstract

In this paper we introduce a voice activity detection (VAD) algorithm that is based on a two-state hidden Markov model. The observation layer of the proposed model, that contains the state conditional probability density functions, is a Gaussian-Laplacian hybrid. The proposed algorithm provides a false detection rate that is significantly lower than that of G. 729 Annex B VAD. Given that it works in the domain of ITU-T G.729 parameters, it requires a minimal additional cost for feature extraction.

Keywords

Gaussian distribution; feature extraction; hidden Markov models; signal detection; statistical analysis; vocoders; Gaussian/Laplacian hybrid; ITU-T G.729; false detection rate; feature extraction; hidden Markov model; probability density; statistical voice activity detector; voice activity detection algorithm; Background noise; Detectors; Frequency; Gaussian noise; Gaussian processes; Hidden Markov models; Laplace equations; Probability density function; Prototypes; Speech enhancement;

fLanguage

English

Publisher

ieee

Conference_Titel

Circuits and Systems, 2003 IEEE 46th Midwest Symposium on

ISSN

1548-3746

Print_ISBN

0-7803-8294-3

Type

conf

DOI

10.1109/MWSCAS.2003.1562381

Filename

1562381