Probability Estimation by Feed-Forward Networks in Continuous Speech Recognition

TitleProbability Estimation by Feed-Forward Networks in Continuous Speech Recognition
Publication TypeTechnical Report
Year of Publication1991
AuthorsRenals, S., Morgan N., & Bourlard H.
Other Numbers660
Abstract

We review the use of feed-forward networks as estimators of probability densities in hidden Markov modeling. In this paper we are mostly concerned with radial basis functions (RBF) networks. We note the isomorphism of RBF networks to tied mixture density estimators; additionally we note that RBF networks are trained to estimate posteriors rather than the likelihoods estimated by tied mixture density estimators. We show how the neural network training should be modified to resolve this mismatch. We also discuss problems with discriminative training, particularly the problem of dealing with unlabeled training data and the mismatch between model and data priors.

URLhttp://www.icsi.berkeley.edu/ftp/global/pub/techreports/1991/tr-91-030.pdf
Bibliographic Notes

ICSI Technical Report TR-91-030

Abbreviated Authors

S. Renals, N. Morgan, and H. Bourlard

ICSI Research Group

Speech

ICSI Publication Type

Technical Report