Originally published at: NVIDIA Research Unveils Flowtron, an Expressive and Natural Speech Synthesis Model | NVIDIA Technical Blog
Many of today’s speech synthesis models lack emotion and human-like expression. To help tackle this problem, a team of researchers from the NVIDIA Applied Deep Learning Research group developed a state-of-the-art model that generates more realistic expressions and provides better user control than previously published models. Named “Flowtron”, the model debuted publicly for the first…