Robust Voice Activity Detection Algorithm for Noisy Speech

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F09%3A00158161" target="_blank" >RIV/68407700:21230/09:00158161 - isvavai.cz</a>

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Robust Voice Activity Detection Algorithm for Noisy Speech

Popis výsledku v původním jazyce

An alternative low-complexity method to identify voice and silence regions of a speech signal is introduced in this paper. The algorithm performance is compared with some other voice classification techniques which are based on energy thresholding. To detect the voice proposed algorithm uses periodicity measure and ratio of high-frequency signal energy to low-frequency signal energy. Total signal energy is also used for voice/silence classification. However, no initial decision about voice detection isbased on the total energy level alone. This makes the algorithm less susceptible to variable acoustic environment. Furthermore, it is possible to use this voice detection scheme for very low-energy speech, which in turn makes the detection more robust insituations where a poor quality microphone is used or where the microphone recording level is low.

Název v anglickém jazyce

Robust Voice Activity Detection Algorithm for Noisy Speech

Popis výsledku anglicky

An alternative low-complexity method to identify voice and silence regions of a speech signal is introduced in this paper. The algorithm performance is compared with some other voice classification techniques which are based on energy thresholding. To detect the voice proposed algorithm uses periodicity measure and ratio of high-frequency signal energy to low-frequency signal energy. Total signal energy is also used for voice/silence classification. However, no initial decision about voice detection isbased on the total energy level alone. This makes the algorithm less susceptible to variable acoustic environment. Furthermore, it is possible to use this voice detection scheme for very low-energy speech, which in turn makes the detection more robust insituations where a poor quality microphone is used or where the microphone recording level is low.

Druh

D - Stať ve sborníku

CEP obor

JA - Elektronika a optoelektronika, elektrotechnika

OECD FORD obor

—

Projekt

—

Návaznosti

Z - Vyzkumny zamer (s odkazem do CEZ)

Rok uplatnění

2009

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

11th International Conference RTT 2009

ISBN

978-80-01-04410-0

ISSN

—

e-ISSN

—

Počet stran výsledku

4

Strana od-do

—

Název nakladatele

České vysoké učení technické v Praze

Místo vydání

Praha

Místo konání akce

Srby

Datum konání akce

2. 9. 2009

Typ akce podle státní příslušnosti

EUR - Evropská akce

Kód UT WoS článku

—