The Effect of a Voice Activity Detector on the Speech Enhancement

Torsten Dau; Jasmina Catic; Jörg Buchholz; Fredrik Gran

doi:10.1155/2010/840294

The Effect of a Voice Activity Detector on the Speech Enhancement

Torsten Dau, Jasmina Catic, Jörg Buchholz, Fredrik Gran

GN ReSound A/S

Research output: Contribution to journal › Journal article › Research › peer-review

461 Downloads (Pure)

Abstract

A multimicrophone speech enhancement algorithm for binaural hearing aids that preserves interaural time delays was proposed recently. The algorithm is based on multichannel Wiener filtering and relies on a voice activity detector (VAD) for estimation of second-order statistics. Here, the effect of a VAD on the speech enhancement of this algorithm was evaluated using an envelopebased VAD, and the performance was compared to that achieved using an ideal error-free VAD. The performance was considered for stationary directional noise and nonstationary diffuse noise interferers at input SNRs from −10 to +5 dB. Intelligibilityweighted SNR improvements of about 20 dB and 6 dB were found for the directional and diffuse noise, respectively. No large degradations (

Original language	English
Journal	Eurasip Journal on Audio, Speech, and Music Processing
ISSN	1687-4714
DOIs	https://doi.org/10.1155/2010/840294
Publication status	Published - 2010

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1155/2010/840294

Catic et al 2010[1]Final published version, 1.42 MB

OpenUrl availability

Full text

Cite this

@article{7aa9bcb19e034f569775930ec35f6333,

title = "The Effect of a Voice Activity Detector on the Speech Enhancement",

abstract = "A multimicrophone speech enhancement algorithm for binaural hearing aids that preserves interaural time delays was proposed recently. The algorithm is based on multichannel Wiener filtering and relies on a voice activity detector (VAD) for estimation of second-order statistics. Here, the effect of a VAD on the speech enhancement of this algorithm was evaluated using an envelopebased VAD, and the performance was compared to that achieved using an ideal error-free VAD. The performance was considered for stationary directional noise and nonstationary diffuse noise interferers at input SNRs from −10 to +5 dB. Intelligibilityweighted SNR improvements of about 20 dB and 6 dB were found for the directional and diffuse noise, respectively. No large degradations (",

author = "Torsten Dau and Jasmina Catic and J{\"o}rg Buchholz and Fredrik Gran",

year = "2010",

doi = "10.1155/2010/840294",

language = "English",

journal = "Eurasip Journal on Audio, Speech, and Music Processing",

issn = "1687-4714",

publisher = "Springer Open",

}

TY - JOUR

T1 - The Effect of a Voice Activity Detector on the Speech Enhancement

AU - Dau, Torsten

AU - Catic, Jasmina

AU - Buchholz, Jörg

AU - Gran, Fredrik

PY - 2010

Y1 - 2010

N2 - A multimicrophone speech enhancement algorithm for binaural hearing aids that preserves interaural time delays was proposed recently. The algorithm is based on multichannel Wiener filtering and relies on a voice activity detector (VAD) for estimation of second-order statistics. Here, the effect of a VAD on the speech enhancement of this algorithm was evaluated using an envelopebased VAD, and the performance was compared to that achieved using an ideal error-free VAD. The performance was considered for stationary directional noise and nonstationary diffuse noise interferers at input SNRs from −10 to +5 dB. Intelligibilityweighted SNR improvements of about 20 dB and 6 dB were found for the directional and diffuse noise, respectively. No large degradations (

AB - A multimicrophone speech enhancement algorithm for binaural hearing aids that preserves interaural time delays was proposed recently. The algorithm is based on multichannel Wiener filtering and relies on a voice activity detector (VAD) for estimation of second-order statistics. Here, the effect of a VAD on the speech enhancement of this algorithm was evaluated using an envelopebased VAD, and the performance was compared to that achieved using an ideal error-free VAD. The performance was considered for stationary directional noise and nonstationary diffuse noise interferers at input SNRs from −10 to +5 dB. Intelligibilityweighted SNR improvements of about 20 dB and 6 dB were found for the directional and diffuse noise, respectively. No large degradations (

U2 - 10.1155/2010/840294

DO - 10.1155/2010/840294

M3 - Journal article

SN - 1687-4714

JO - Eurasip Journal on Audio, Speech, and Music Processing

JF - Eurasip Journal on Audio, Speech, and Music Processing

ER -

The Effect of a Voice Activity Detector on the Speech Enhancement

Abstract

UN SDGs

Access to Document

OpenUrl availability

Fingerprint

Cite this