Feature-based pronunciation modeling for speech recognition

We present an approach to pronunciation mod-eling in which the evolution of multiple lin-guistic feature streams is explicitly represented. This differs from phone-based models in that pronunciation variation is viewed as the result of feature asynchrony and changes in feature values, rather than phone substitutions, inser-tions, and deletions. We have implemented a flexible feature-based pronunciation model us-ing dynamic Bayesian networks. In this paper, we describe our approach and report on a pilot experiment using phonetic transcriptions of ut-terances from the Switchboard corpus. The ex-perimental results, as well as the model’s quali-tative behavior, suggest that this is a promising way of accounting for the types of pronuncia-tion variation often seen in spontaneous speech.