Document-Level Neural TTS Using Curriculum Learning and Attention Masking

Document-Level Neural TTS Using Curriculum Learning and Attention Masking

Speech synthesis has been developed to the level of natural human-level speech synthesized through an attention-based end-to-end text-to-speech synthesis (TTS) model. However, it is difficult to generate attention when synthesizing a text longer than the trained length or document-level text. In thi...

Full description

Bibliographic Details
Main Authors: Sung-Woong Hwang, Joon-Hyuk Chang
Format: Article
Language:English
Published: IEEE 2021-01-01
Series:IEEE Access
Subjects:
Speech synthesis
document-level neural TTS
curriculum learning
attention masking
Tacotron2
MelGAN
Online Access:https://ieeexplore.ieee.org/document/9312676/

Internet

https://ieeexplore.ieee.org/document/9312676/

Similar Items