Publication Details
The Spoken Web Search Task
Metze Florian (CMU)
Buzo Andi (UPB)
Szőke Igor, Ing., Ph.D. (DCGM FIT BUT)
Rodriguez-Fuentes Luis J. (EHU)
keyword search, query by example, spoken term detection
The definition of Spoken Web Search Task for low and zero resourced languages which was held as a part of MediaEval campaign 2013. The purpose of the task is to search audio content using audio query. The data setup contains 20 hours of data - mix of 9 languages a several acoustic conditions. The scoring metrics are TWV and NCE.
In this paper, we describe the "Spoken Web Search" Task, which is being held as part of the 2013 MediaEval campaign. The purpose of this task is to perform audio search in multiple languages and acoustic conditions, with very few resources being available for each individual language. This year the data contains audio from nine different languages and is much bigger in size than in previous years, mimicking realistic low/zero - resource settings.
@INPROCEEDINGS{FITPUB10503, author = "Xavier Anguera and Florian Metze and Andi Buzo and Igor Sz\H{o}ke and J. Luis Rodriguez-Fuentes", title = "The Spoken Web Search Task", pages = "1--2", booktitle = "CEUR Workshop Proceedings", journal = "CEUR Workshop Proceedings", volume = 2013, number = 1043, year = 2013, location = "Barcelona, ES", publisher = "CEUR-WS.org", ISSN = "1613-0073", language = "english", url = "https://www.fit.vut.cz/research/publication/10503" }