Fast and High-Quality Singing Voice Synthesis System based on Convolutional Neural Networks

Presentation and demos of our ICASSP 2020 paper. “Fast and High-Quality Singing Voice Synthesis System based on Convolutional Neural Networks“ Kazuhiro Nakamura 1, Shinji Takaki 1,2, Kei Hashimoto 1,2, Keiichiro Oura 1,2, Yoshihiko Nankaku 2, and Keiichi Tokuda 1,2 1: Techno-Speech, Inc. 2: Nagoya Institute of Technology Paper: Samples: