Článek ve sborníku konference

 
Tejedor, J., Szőke, I., Fapšo, M.: Novel Methods for Query Selection and Query Combination in Query-By-Example Spoken Term Detection, In: Proceedings of the ACM Multimedia 2010 International Conference, Florencie, IT, ACM, 2010, s. 15-20, ISBN 978-1-60558-933-6
Jazyk publikace:angličtina
Název publikace:Novel Methods for Query Selection and Query Combination in Query-By-Example Spoken Term Detection
Název (cs):Nové přístupy pro výběr a kombinaci vzoru klíčových slov v detekci klíčových slov zadaných vzorem
Strany:15-20
Sborník:Proceedings of the ACM Multimedia 2010 International Conference
Konference:ACM Multimedia 2010 International Conference
Řada knih:Copyright 2010 ACM 978-1-4503-0162-6/10/10
Místo vydání:Florencie, IT
Rok:2010
ISBN:978-1-60558-933-6
Vydavatel:Association for Computing Machinery
URL:http://www.fit.vutbr.cz/research/groups/speech/publi/2010/tejedor_szoke_SSCS_2010_p15.pdf [PDF]
Klíčová slova
query-by-example,query selection,query combination,speech recognition
Anotace
Článek pojednává o nových přístupech pro výběr a kombinaci vzoru klíčových slov v detekci klíčových slov zadaných vzorem.
Abstrakt
Query-by-example (QbE) spoken term detection (STD) is necessary for low-resource scenarios where training material is hardly available and word-based speech recognition systems cannot be employed. We present two novel contributions to QbE STD: the first introduces several criteria to select the optimal example used as query throughout the search system. The second presents a novel feature level example combination to construct a more robust query used during the search. Experiments, tested on with-in language and cross-lingual QbE STD setups, show a significant improvement when the query is selected according to an optimal criterion over when the query is selected randomly for both setups and a significant improvement when several examples are combined to build the input query for the search system compared with the use of the single best example. They also show comparable performance to that of a stateof- the-art acoustic keyword spotting system.
BibTeX:
@INPROCEEDINGS{
   author = {Javier Tejedor and Igor Szőke and Michal Fapšo},
   title = {Novel Methods for Query Selection and Query Combination in
	Query-By-Example Spoken Term Detection},
   pages = {15--20},
   booktitle = {Proceedings of the ACM Multimedia 2010 International
	Conference},
   series = {Copyright 2010 ACM 978-1-4503-0162-6/10/10},
   year = {2010},
   location = {Florencie, IT},
   publisher = {Association for Computing Machinery},
   ISBN = {978-1-60558-933-6},
   language = {english},
   url = {http://www.fit.vutbr.cz/research/view_pub.php?id=9461}
}