Synchronising speech segments with musical beats in Mandarin and English singing

Cong Zhang, Jian Zhu

Generating synthesised singing voice with models trained on speech data has many advantages due to the models' flexibility and controllability. However, since the information about…