Generative Model-Based Text-to-Speech Synthesis
Abstract
Recent progress in generative modeling has improved the naturalness of synthesized speech significantly. In this talk I will summarize these generative model-based approaches for speech synthesis and describe possible future directions.