公开(公告)号：US10510358B1

公开(公告)日：2019-12-17

申请号：US15719950

申请日：2017-09-29

Applicant: Amazon Technologies, Inc.

Inventor： Roberto Barra-Chicote ,  Alexis Moinet ,  Nikko Strom

IPC: G10L21/038 ,  G10L19/07 ,  G10L21/02 ,  G10L13/08 ,  G10L25/30 ,  G10L13/047

Abstract: An approach to speech synthesis uses two phases in which a relatively low quality waveform is computed, and that waveform is passed through an enhancement phase which generates the waveform that is ultimately used to produce the acoustic signal provided to the user. For example, the first phase and the second phase are each implemented using a separate artificial neural network. The two phases may be computationally preferable to using a direct approach to yield a synthesized waveform of comparable quality.