ISCA Archive Interspeech 2006
ISCA Archive Interspeech 2006

Issues with uncertainty decoding for noise robust speech recognition

H. Liao, M. J. F. Gales

Recently there has been interest in uncertainty decoding for robust speech recognition. Here the uncertainty associated with the observation in noise is propagated to the recogniser. By using appropriate approximations for this uncertainty, it is possible to obtain efficient implementations during decoding. The aim of these schemes is to obtain performance which is close to that of a model-based compensated system, without the computational cost. Unfortunately, in low SNR there is a fundamental issue with front-end uncertainty decoding where the model means and variances are updated according to the features. This is described in detail using the Joint and SPLICE with uncertainty forms, but is not limited to these two techniques. A solution for the Joint scheme is presented along with the implicit approach used in SPLICE with uncertainty. In addition, a model-based Joint uncertainty scheme is described, which is more efficient and powerful than the front-end schemes, and being model-based not affected by this problem. This issue is illustrated using the AURORA 2.0 database with these various systems.

doi: 10.21437/Interspeech.2006-343

Cite as: Liao, H., Gales, M.J.F. (2006) Issues with uncertainty decoding for noise robust speech recognition. Proc. Interspeech 2006, paper 1627-Tue2BuP.8, doi: 10.21437/Interspeech.2006-343

  author={H. Liao and M. J. F. Gales},
  title={{Issues with uncertainty decoding for noise robust speech recognition}},
  booktitle={Proc. Interspeech 2006},
  pages={paper 1627-Tue2BuP.8},