DocumentCode
675468
Title
Speech resources for a Serbian LVCSR system
Author
Ostrogonac, Stevan ; Suzic, Sinisa ; Bojanic, Milana ; Pakoci, Edvin
Author_Institution
Fac. of Tech. Sci., Univ. of Novi Sad, Novi Sad, Serbia
fYear
2013
fDate
26-28 Nov. 2013
Firstpage
478
Lastpage
481
Abstract
This paper describes the whole procedure of speech database collection and processing required for building a good large vocabulary speech recognition system for the Serbian language. The speech database consists of speech recordings from audio books, radio programs and talk shows, as well as read utterances from an array of male and female speakers. To date, around 200 hours of read speech is collected, as well as about 10 hours of radio recordings.
Keywords
natural language processing; speech recognition; Serbian LVCSR system; audio books; large vocabulary speech recognition system; radio programs; radio recording; speech database collection; speech recording; speech resources; talk show; Acoustics; Databases; Materials; Speech; Speech recognition; Training; Vocabulary; Serbian; large vocabulary continuous speech recognition; speech database;
fLanguage
English
Publisher
ieee
Conference_Titel
Telecommunications Forum (TELFOR), 2013 21st
Conference_Location
Belgrade
Print_ISBN
978-1-4799-1419-7
Type
conf
DOI
10.1109/TELFOR.2013.6716271
Filename
6716271
Link To Document