i-ETTS

Reinforcement Learning for Emotional Text-to-Speech Synthesis with Improved Emotion Discriminability

Authors: Rui Liu, Berrak Sisman and Haizhou Li

This paper was accepted by InterSpeech2021.

Note: To our best knowledge, this is the first study of reinforcement learning in emotional text-to-speech synthesis.

Speech samples

Speech samples are available at demo page.

Citing

To cite this repository:

@inproceedings{liu21p_interspeech,
  title     = {Reinforcement Learning for Emotional Text-to-Speech Synthesis with Improved Emotion Discriminability},
  author    = {Rui Liu and Berrak Sisman and Haizhou Li},
  year      = {2021},
  booktitle = {Interspeech 2021},
  pages     = {4648--4652},
  doi       = {10.21437/Interspeech.2021-1236},
  issn      = {2958-1796},
}