Fetching the paper…

Style Tokens: Unsupervised Style Modeling, Control and Transfer in End-to-End Speech Synthesis · Around