prophesier/diff-svc
A diffusion-based singing voice conversion system that transforms input singing voices into a target timbre with pitch correction support.
Not currently ranked — collecting fresh signals.
star history
This repository implements singing voice conversion using diffusion models. It converts input singing voices into a target timbre while supporting basic pitch correction. The system uses deep learning components including HuBERT and ContentVec for content feature extraction, along with neural vocoders for audio synthesis. The project supports training on custom datasets and provides GPU-accelerated inference for voice conversion tasks.
Frequently asked
- What is prophesier/diff-svc?
- A diffusion-based singing voice conversion system that transforms input singing voices into a target timbre with pitch correction support.
- Is diff-svc open source?
- Yes — prophesier/diff-svc is open source, released under the AGPL-3.0 license.
- What language is diff-svc written in?
- prophesier/diff-svc is primarily written in Jupyter Notebook.
- How popular is diff-svc?
- prophesier/diff-svc has 2.7k stars on GitHub.
- Where can I find diff-svc?
- prophesier/diff-svc is on GitHub at https://github.com/prophesier/diff-svc.