Improving Emotional Speech Synthesis by Using SUS-Constrained VAE and Text Encoder Aggregation.

Fengyu Yang Jian Luan Yujun Wang

Published in: ICASSP (2022)

Keyphrases

speech synthesis
text to speech
speech recognition
prosodic features
vocal tract
information retrieval
text mining
text retrieval
speech corpus
neural network
bit rate
search engine
database
motion estimation
pattern recognition
non stationary
word processing