Voice activity detection in noisy environments based on double-combined fourier transform and line fitting

Jinsoo Park, Wooil Kim, David K. Han, Hanseok Ko

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

A new voice activity detector for noisy environments is proposed. In conventional algorithms, the endpoint of speech is found by applying an edge detection filter that finds the abrupt changing point in a feature domain. However, since the frame energy feature is unstable in noisy environments, it is difficult to accurately find the endpoint of speech. Therefore, a novel feature extraction algorithm based on the double-combined Fourier transform and envelope line fitting is proposed. It is combined with an edge detection filter for effective detection of endpoints. Effectiveness of the proposed algorithm is evaluated and compared to other VAD algorithms using two different databases, which are AURORA 2.0 database and SITEC database. Experimental results show that the proposed algorithm performs well under a variety of noisy conditions.

Original languageEnglish
Pages (from-to)146040
Number of pages1
JournalTheScientificWorldJournal
Volume2014
DOIs
Publication statusPublished - 2014

Fingerprint

Fourier Analysis
Fourier transform
Fourier transforms
Edge detection
Databases
filter
Feature extraction
detection
Detectors
energy
speech

ASJC Scopus subject areas

  • Medicine(all)

Cite this

Voice activity detection in noisy environments based on double-combined fourier transform and line fitting. / Park, Jinsoo; Kim, Wooil; Han, David K.; Ko, Hanseok.

In: TheScientificWorldJournal, Vol. 2014, 2014, p. 146040.

Research output: Contribution to journalArticle

@article{93019d8d747f4ac7bb2f39ffc4ce4b05,
title = "Voice activity detection in noisy environments based on double-combined fourier transform and line fitting",
abstract = "A new voice activity detector for noisy environments is proposed. In conventional algorithms, the endpoint of speech is found by applying an edge detection filter that finds the abrupt changing point in a feature domain. However, since the frame energy feature is unstable in noisy environments, it is difficult to accurately find the endpoint of speech. Therefore, a novel feature extraction algorithm based on the double-combined Fourier transform and envelope line fitting is proposed. It is combined with an edge detection filter for effective detection of endpoints. Effectiveness of the proposed algorithm is evaluated and compared to other VAD algorithms using two different databases, which are AURORA 2.0 database and SITEC database. Experimental results show that the proposed algorithm performs well under a variety of noisy conditions.",
author = "Jinsoo Park and Wooil Kim and Han, {David K.} and Hanseok Ko",
year = "2014",
doi = "10.1155/2014/146040",
language = "English",
volume = "2014",
pages = "146040",
journal = "Scientific World Journal",
issn = "2356-6140",
publisher = "Hindawi Publishing Corporation",

}

TY - JOUR

T1 - Voice activity detection in noisy environments based on double-combined fourier transform and line fitting

AU - Park, Jinsoo

AU - Kim, Wooil

AU - Han, David K.

AU - Ko, Hanseok

PY - 2014

Y1 - 2014

N2 - A new voice activity detector for noisy environments is proposed. In conventional algorithms, the endpoint of speech is found by applying an edge detection filter that finds the abrupt changing point in a feature domain. However, since the frame energy feature is unstable in noisy environments, it is difficult to accurately find the endpoint of speech. Therefore, a novel feature extraction algorithm based on the double-combined Fourier transform and envelope line fitting is proposed. It is combined with an edge detection filter for effective detection of endpoints. Effectiveness of the proposed algorithm is evaluated and compared to other VAD algorithms using two different databases, which are AURORA 2.0 database and SITEC database. Experimental results show that the proposed algorithm performs well under a variety of noisy conditions.

AB - A new voice activity detector for noisy environments is proposed. In conventional algorithms, the endpoint of speech is found by applying an edge detection filter that finds the abrupt changing point in a feature domain. However, since the frame energy feature is unstable in noisy environments, it is difficult to accurately find the endpoint of speech. Therefore, a novel feature extraction algorithm based on the double-combined Fourier transform and envelope line fitting is proposed. It is combined with an edge detection filter for effective detection of endpoints. Effectiveness of the proposed algorithm is evaluated and compared to other VAD algorithms using two different databases, which are AURORA 2.0 database and SITEC database. Experimental results show that the proposed algorithm performs well under a variety of noisy conditions.

UR - http://www.scopus.com/inward/record.url?scp=84929938861&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84929938861&partnerID=8YFLogxK

U2 - 10.1155/2014/146040

DO - 10.1155/2014/146040

M3 - Article

C2 - 25170520

AN - SCOPUS:84929938861

VL - 2014

SP - 146040

JO - Scientific World Journal

JF - Scientific World Journal

SN - 2356-6140

ER -