This AI Paper Proposes CoMoSVC: A Consistency Model-based SVC Method that Aims to Achieve both High-Quality Generation and High-Speed Sampling

CoMoSVC is an innovative method for singing voice conversion that aims to achieve high-quality generation and high-speed sampling. It leverages the consistency model and a teacher-student model framework to provide rapid and high-quality voice conversion. CoMoSVC outperforms traditional diffusion-based SVC systems in terms of Show more