NVIDIA Research Unveils Flowtron, an Expressive and Natural Speech Synthesis Model
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.