Memory efficient modeling of polyphone context with weighted finite-state transducers.

Emilian Stoimenov, John W. McDonough

Browse the full Interspeech paper archive.