Pith. sign in

REVIEW

DiffSVC: A Diffusion Probabilistic Model for Singing Voice Conversion

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2105.13871 v1 pith:4B4JRAXF submitted 2021-05-28 eess.AS cs.CLcs.SD

DiffSVC: A Diffusion Probabilistic Model for Singing Voice Conversion

classification eess.AS cs.CLcs.SD
keywords diffsvcvoiceconversiondenoisingdiffusionfeaturessinginginput
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

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.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.