Combined Acoustic and Pronunciation Modelling for Non-Native Speech Recognition - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2007

Combined Acoustic and Pronunciation Modelling for Non-Native Speech Recognition

Dominique Fohr
Irina Illina

Abstract

In this paper, we present several adaptation methods for non-native speech recognition. We have tested pronunciation modelling, MLLR and MAP non-native pronunciation adaptation and HMM models retraining on the HIWIRE foreign accented English speech database. The ``phonetic confusion'' scheme we have developed consists in associating to each spoken phone several sequences of confused phones. In our experiments, we have used different combinations of acoustic models representing the canonical and the foreign pronunciations: spoken and native models, models adapted to the non-native accent with MAP and MLLR. The joint use of pronunciation modelling and acoustic adaptation led to further improvements in recognition accuracy. The best combination of the above mentioned techniques resulted in a relative word error reduction ranging from 46% to 71%.
Fichier principal
Vignette du fichier
is2007_latex_template.pdf (81.83 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

inria-00184560 , version 1 (05-11-2007)

Identifiers

  • HAL Id : inria-00184560 , version 1
  • ARXIV : 0711.0811

Cite

Ghazi Bouselmi, Dominique Fohr, Irina Illina. Combined Acoustic and Pronunciation Modelling for Non-Native Speech Recognition. InterSpeech 2007, Aug 2007, Antwerp, Belgium. ⟨inria-00184560⟩
88 View
227 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More