Publication Details

Acoustic keyword spotter - optimization from end-user perspective

SZŐKE Igor, GRÉZL František, ČERNOCKÝ Jan and FAPŠO Michal. Acoustic keyword spotter - optimization from end-user perspective. In: Proceedings of the 2010 IEEE Spoken Language Technology Workshop. IEEE Catalog Number: CFP 10SLT-USB. Berkeley, California: IEEE Signal Processing Society, 2010, pp. 177-181. ISBN 978-1-4244-7902-3.
Czech title
Akustický vyhledávač klíčových slov -optimalizace z perspektivy koncového uživatele
Type
conference paper
Language
english
Authors
Szőke Igor, Ing., Ph.D. (DCGM FIT BUT)
Grézl František, Ing., Ph.D. (DCGM FIT BUT)
Černocký Jan, prof. Dr. Ing. (DCGM FIT BUT)
Fapšo Michal, Ing. (DCGM FIT BUT)
URL
Keywords

keyword spotting, spoken term detection, neural networks, calibration

Abstract

This paper is on acoustic keyword spotting. It presents several steps that have to be done to obtain a usable acoustic keyword spotting system. The novelty of the system is in the calibration.

Annotation

The paper deals with the development of acoustic keyword spotter (KWS) meeting requirements of a real user from the security community. While the basic scheme of the KWS is relatively standard, it uses novel features derived by a hierarchy of neural networks, and score normalization trained to maximize a user-like evaluation metric. The results are reported on a selection of Czech conversational telephone speech (CTS), radio and read data.

Published
2010
Pages
177-181
Proceedings
Proceedings of the 2010 IEEE Spoken Language Technology Workshop
Series
IEEE Catalog Number: CFP 10SLT-USB
Conference
IEEE Workshop on Spoken Language Technology, Berkeley, US
ISBN
978-1-4244-7902-3
Publisher
IEEE Signal Processing Society
Place
Berkeley, California, US
BibTeX 
@INPROCEEDINGS{FITPUB9457,
   author = "Igor Sz\H{o}ke and Franti\v{s}ek Gr\'{e}zl and Jan \v{C}ernock\'{y} and Michal Fap\v{s}o",
   title = "Acoustic keyword spotter - optimization from end-user perspective",
   pages = "177--181",
   booktitle = "Proceedings of the 2010 IEEE Spoken Language Technology Workshop",
   series = "IEEE Catalog Number: CFP 10SLT-USB",
   year = 2010,
   location = "Berkeley, California, US",
   publisher = "IEEE Signal Processing Society",
   ISBN = "978-1-4244-7902-3",
   language = "english",
   url = "https://www.fit.vut.cz/research/publication/9457"
}
Back to top