A trainable excitation model for HMM-based speech synthesis.

Ranniery Maia, Tomoki Toda, Heiga Zen, Yoshihiko Nankaku, Keiichi Tokuda

Browse the full Interspeech paper archive.