Statistical singing voice conversion with direct waveform modification based on the spectrum differential

Kazuhiro Kobayashi, Tomoki Toda, Graham Neubig, Sakriani Sakti, Satoshi Nakamura