Publication Details

Spoken Term Detection System Based on a Combination of LVCSR and Phonetic Search

SZŐKE Igor, FAPŠO Michal, KARAFIÁT Martin, BURGET Lukáš, GRÉZL František, SCHWARZ Petr, GLEMBEK Ondřej, MATĚJKA Pavel, KOPECKÝ Jiří and ČERNOCKÝ Jan. Spoken Term Detection System Based on a Combination of LVCSR and Phonetic Search. Brno, 2007.
Czech title
Detekce dotazů v mluvených dokumentech založená na kombinaci LVCSR a fonetického vyhledávání
Type
presentation,poster
Language
english
Authors
URL
Keywords

spoken term detection

Abstract

The paper describes Spoken Term Detection System Based on a Combination of LVCSR and Phonetic Search.

Annotation

The paper presents Brno University of Technology (BUT) system for indexing and search of speech, combining LVCSR and phonetic approach. It brings a complete description of individual building blocks of the system from signal processing, through the recognizers, indexing and search until the normalization of detection scores. It also describes the data used in the first edition of NIST Spoken term detection (STD) evaluation. The results are presented on three US-English conditions - meetings, broadcast news and conversational telephone speech, in terms of detection error trade-off (DET) curves and term-weighted values (TWV) metrics defined by NIST.

Published
2007
Pages
1-1
Conference
Machine Learning and Multimodal Interaction, Brno, CZ
Place
Brno, CZ
Back to top