ESCA Workshop on Prosody Lund, Sweden September 27-29, 1993

Lexicon and Prosodic Structure in a Text-To-Speech System

Sverre Stensby, Berit Horvei, Georg E. Ottesen

SINTEF DELAB, Trondheim, Norway

This article is a presentation of a pronunciation lexicon and its use in improving prosody in a text-to-speech system. The inclusion of a lexicon is a natural way of giving synthesis systems information of word class and word pronunciation. The grammatical information is used to disambiguate homographs and to form a basis for prosodic structuring. The realized prosody is based on patterns for rhythmical units named feet and utilizes the information of word tone and pronunciation found in the lexicon.

Full Paper

Bibliographic reference.  Stensby, Sverre / Horvei, Berit / Ottesen, Georg E. (1993): "Lexicon and prosodic structure in a text-to-speech system", In Prosody-1993, 82-85.