logo Idiap Research Institute        
 [BibTeX] [Marc21]
Using out-of-language data to improve an under-resourced speech recognizer
Type of publication: Idiap-RR
Citation: Imseng_Idiap-RR-09-2013
Number: Idiap-RR-09-2013
Year: 2013
Month: 3
Institution: Idiap
Abstract: Under-resourced speech recognizers may benefit from data in languages other than the target language. In this paper, we report how to boost the performance of an Afrikaans automatic speech recognition system by using already available Dutch data. We successfully exploit available multilingual resources through 1) posterior features, estimated by multilayer perceptrons (MLP) and 2) subspace Gaussian mixture models (SGMMs). Both the MLPs and the SGMMs can be trained on out-of-language data. We use three different acoustic modeling techniques, namely Tandem, Kullback-Leibler divergence based HMMs (KL-HMM) as well as SGMMs and show that the proposed multilingual systems yield 12% relative improvement compared to a conventional monolingual HMM/GMM system only trained on Afrikaans. We also show that KL-HMMs are extremely powerful for under-resourced languages: using only six minutes of Afrikaans data (in combination with out-of-language data), KL-HMM yields about 30% relative improvement compared to conventional maximum likelihood linear regression and maximum a posteriori based acoustic model adaptation.
Crossref: Imseng_SPECOM_2013:
Projects Idiap
Authors Imseng, David
Motlicek, Petr
Bourlard, Hervé
Garner, Philip N.
Added by: [ADM]
Total mark: 0
  • Imseng_Idiap-RR-09-2013.pdf (MD5: d954339f771651dfabbd2f8576dfdf3d)