Representation of large lexica using finite-state transducers for the multilingual text-to-speech synthesis systems.

Matej Rojc, Zdravko Kacic

Browse the full Interspeech paper archive.