Transforming Spectrum and Prosody for Emotional Voice Conversion with Non-Parallel Training Data
2020·Arxiv