Selection of an optimum speech parameterization for continuous speech recognition system using a telephone channel

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F49777513%3A23520%2F01%3A00065619" target="_blank" >RIV/49777513:23520/01:00065619 - isvavai.cz</a>

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Selection of an optimum speech parameterization for continuous speech recognition system using a telephone channel

Popis výsledku v původním jazyce

This paper describes our effort to select an optimum speech parameterization used in a continuous speech recognition module developed at the Department of Cybernetics,University of West Bohemia,that works with speech of a telephone quality. To design anoptimum front-end module of the mentioned telephone Dialogue system we tested and compared two parameterization techniques with many modifications that could alleviate different properties of telephone channels and assure a high quality of recognition and robustness of a recognizer. The quality of the parameterization and the front-end module was evaluated through the WER during a large number of experiments with recognition of speaker independent continuous speech.

Název v anglickém jazyce

Selection of an optimum speech parameterization for continuous speech recognition system using a telephone channel

Popis výsledku anglicky

This paper describes our effort to select an optimum speech parameterization used in a continuous speech recognition module developed at the Department of Cybernetics,University of West Bohemia,that works with speech of a telephone quality. To design anoptimum front-end module of the mentioned telephone Dialogue system we tested and compared two parameterization techniques with many modifications that could alleviate different properties of telephone channels and assure a high quality of recognition and robustness of a recognizer. The quality of the parameterization and the front-end module was evaluated through the WER during a large number of experiments with recognition of speaker independent continuous speech.

Druh

D - Stať ve sborníku

CEP obor

JD - Využití počítačů, robotika a její aplikace

OECD FORD obor

—

Projekt

—

Návaznosti

Z - Vyzkumny zamer (s odkazem do CEZ)

Rok uplatnění

2001

Kód důvěrnosti údajů

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

Název statě ve sborníku

Selection of an optimum speech parameterization for continuous speech recognition system using a telephone channel

ISBN

—

ISSN

—

e-ISSN

—

Počet stran výsledku

4

Strana od-do

—

Název nakladatele

Neuveden

Místo vydání

Neuveden

Místo konání akce

Neuveden

Datum konání akce

1. 1. 2001

Typ akce podle státní příslušnosti

CST - Celostátní akce

Kód UT WoS článku

—