DiffSVC: A Diffusion Probabilistic Model for Singing Voice Conversion
This work addresses the problem of generating high-fidelity and expressive singing voices for human-computer interaction, representing an incremental improvement over existing methods.
The paper tackles singing voice conversion by proposing DiffSVC, a system based on a diffusion probabilistic model that uses phonetic posteriorgrams and other features to achieve superior performance in naturalness and voice similarity compared to state-of-the-art approaches.
Singing voice conversion (SVC) is one promising technique which can enrich the way of human-computer interaction by endowing a computer the ability to produce high-fidelity and expressive singing voice. In this paper, we propose DiffSVC, an SVC system based on denoising diffusion probabilistic model. DiffSVC uses phonetic posteriorgrams (PPGs) as content features. A denoising module is trained in DiffSVC, which takes destroyed mel spectrogram produced by the diffusion/forward process and its corresponding step information as input to predict the added Gaussian noise. We use PPGs, fundamental frequency features and loudness features as auxiliary input to assist the denoising process. Experiments show that DiffSVC can achieve superior conversion performance in terms of naturalness and voice similarity to current state-of-the-art SVC approaches.