Lightly supervised discriminative training of grapheme models for improved sentence-level alignment of speech and text data.

Adriana Stan, Peter Bell, Junichi Yamagishi, Simon King

Browse the full Interspeech paper archive.