logo Idiap Research Institute        
 [BibTeX] [Marc21]
Development of Bilingual ASR System for MediaParl Corpus
Type of publication: Idiap-RR
Citation: Motlicek_Idiap-RR-21-2014
Number: Idiap-RR-21-2014
Year: 2014
Month: 12
Institution: Idiap
Address: Rue Marconi 19
Abstract: The development of an Automatic Speech Recognition (ASR) system for the bilingual MediaParl corpus is challenging for several reasons: (1) reverberant recordings, (2) accented speech, and (3) no prior information about the language. In that context, we employ frequency domain linear prediction-based (FDLP) features to reduce the effect of reverberation, exploit bilingual deep neural networks applied in Tandem and hybrid acoustic modeling approaches to significantly improve ASR for accented speech and develop a fully bilingual ASR system using entropy-based decoding-graph selection. Our experiments indicate that the proposed bilingual ASR system performs similar to a language-specific ASR system if approximately five seconds of speech are available.
Keywords: lan- guage identification, Multilingual automatic speech recognition, non-native speech
Projects Idiap
SAMSUNG
DBOX
Authors Motlicek, Petr
Imseng, David
Cernak, Milos
Kim, Namhoon
Crossref by Motlicek_INTERSPEECH2014_2014
Added by: [ADM]
Total mark: 0
Attachments
  • Motlicek_Idiap-RR-21-2014.pdf (MD5: 9e2d1f1b9aede4d255d13482850c8e04)
Notes