PlayVoice/whisper-vits-svc
A deep learning model for end-to-end singing voice conversion using VITS (Variational Inference with adversarial learning).
Not currently ranked — collecting fresh signals.
star history
This project implements a variational inference model with adversarial learning for singing voice conversion based on the VITS architecture. It enables converting one singer’s voice to another speaker’s voice, supports multiple speakers, speaker mixing, and basic F0 editing. The model requires a minimum of 6GB VRAM for training and can even handle audio with light accompaniment.
Frequently asked
- What is PlayVoice/whisper-vits-svc?
- A deep learning model for end-to-end singing voice conversion using VITS (Variational Inference with adversarial learning).
- Is whisper-vits-svc open source?
- Yes — PlayVoice/whisper-vits-svc is open source, released under the MIT license.
- What language is whisper-vits-svc written in?
- PlayVoice/whisper-vits-svc is primarily written in Python.
- How popular is whisper-vits-svc?
- PlayVoice/whisper-vits-svc has 2.9k stars on GitHub.
- Where can I find whisper-vits-svc?
- PlayVoice/whisper-vits-svc is on GitHub at https://github.com/PlayVoice/whisper-vits-svc.