NNSVS NNSVS is a neural network based singing voice synthesis library that supports multiple languages and research-friendly features. While it supports multiple languages, most songs found on the internet are in Japanese.
The original version of Sinsy is a HMM-based singing voice synthesis system, which only supports the Japanese language. Today Sinsy support many other languages using eSpeak as a frontend. QTAU also uses the MusicXML code from Sinsy.