Investigating on Incorporating Pretrained and Learnable Speaker Representations for Multi-Speaker Multi-Style Text-to-Speech

Published in IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2021