Skip to content

Publications

Detail of publication

Citation

Josef V. Psutka and Jan Vaněk and Josef Psutka : Speaker-clustered Acoustic Models Evaluated on GPU for on-line Subtitling of Parliament Meetings . Text, Speech, and Dialogue, Lecture Notes in Computer Science, vol. 6836, p. 284-290, Springer, 2011.

Download PDF

PDF

Abstract

This paper describes the effort with building speaker-clustered acoustic models as a part of the real-time LVCSR system that is used more than one year by the Czech TV for automatic subtitling of parliament meetings broadcasted on the channel ČT24. Speaker-clustered acoustic models are more acoustically homogeneous and therefore give better recognition performance than single gender-independent model or even gender-dependent models. Frequent changes of speakers and a direct connection of the LVCSR system to the audio channel require an automatic switching/fusion of models as quickly as possible. An important part of the solution is real time likelihood evaluations of all clustered acoustic models, taking advantage of a fast GPU(Graphic Processing Unit). The proposed method achieved a WER reduction to the baseline gender-independent model over 2.34% relatively with more than 2M Gaussian mixtures evaluated in real-time.

Detail of publication

Title: Speaker-clustered Acoustic Models Evaluated on GPU for on-line Subtitling of Parliament Meetings
Author: Josef V. Psutka ; Jan Vaněk ; Josef Psutka
Language: English
Date of publication: 6 Sep 2011
Year: 2011
Type of publication: Papers in proceedings of reviewed conferences
Book title: Text, Speech, and Dialogue
Series: Lecture Notes in Computer Science
Číslo vydání: 6836
Page: 284 - 290
DOI: 10.1007/978-3-642-23538-2_36
ISBN: 978-3-642-23537-5
Publisher: Springer
/ 2014-11-12 12:26:51 /

BibTeX

@INPROCEEDINGS{JosefVPsutka_2011_Speaker-clustered,
 author = {Josef V. Psutka and Jan Van\v{e}k and Josef Psutka},
 title = {Speaker-clustered Acoustic Models Evaluated on GPU for on-line Subtitling of Parliament Meetings },
 year = {2011},
 publisher = {Springer},
 volume = {6836},
 pages = {284-290},
 booktitle = {Text, Speech, and Dialogue},
 series = {Lecture Notes in Computer Science},
 ISBN = {978-3-642-23537-5},
 doi = {10.1007/978-3-642-23538-2_36},
 url = {http://www.kky.zcu.cz/en/publications/JosefVPsutka_2011_Speaker-clustered},
}