Experimental Setup of LVCSR System Based on HTK Tools for Evaluation and Development Purposes

Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F12%3A00196167" target="_blank" >RIV/68407700:21230/12:00196167 - isvavai.cz</a>
Result on the web
—
DOI - Digital Object Identifier
—

Result language
angličtina
Original language name
Experimental Setup of LVCSR System Based on HTK Tools for Evaluation and Development Purposes
Original language description
This paper deals with construction of a large vocabulary continuous speech recognition (LVCSR) system for Czech language, using publicly available tools, which can serve mainly for development purposes. The main focus is on the acoustic model creation suitable for further use in the LVCSR system decoding, taking into account the limitations of the used decoder, one of which is the need to create the cross-word triphones. For experimental setup tree-based clustering algorithm was chosen for state clustering. Signals for the training and testing purposes come from the SPEECON database and were selected based on their low SNR. All these tasks were done with the help of the HTK Toolkit and PERL programming language.
Czech name
—
Czech description
—

Publication year
2012
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Article name in the collection
POSTER 2012 - 16th International Student Conference on Electrical Engineering
ISBN
978-80-01-05043-9
ISSN
—
e-ISSN
—
Number of pages
4
Pages from-to
1-4
Publisher name
Czech Technical University in Prague
Place of publication
Praha
Event location
Prague
Event date
May 17, 2012
Type of event by nationality
EUR - Evropská akce
UT code for WoS article
—

Similar results(10)