Using deep bidirectional recurrent neural networks for prosodic-target prediction in a unit-selection text-to-speech system.

Raul Fernandez, Asaf Rendel, Bhuvana Ramabhadran, Ron Hoory

Browse the full Interspeech paper archive.