Text-to-speech synthesis using spectral modeling based on non-negative autoencoder.

Takeru Gorai, Daisuke Saito, Nobuaki Minematsu

Browse the full Interspeech paper archive.