mchong6/JoJoGAN
A one-shot face stylization model that fine-tunes StyleGAN on a single reference image to apply artistic styles to any face.
Not currently ranked — collecting fresh signals.
star history
JoJoGAN is a one-shot image stylization system that learns to transfer artistic styles (such as anime) from a single reference image onto any input face. It approximates paired training data through GAN inversion and fine-tunes a pretrained StyleGAN to capture stylistic details like eye shapes and line boldness. The model enables users to stylize images via Google Colab, Replicate, or Hugging Face Spaces.
Frequently asked
- What is mchong6/JoJoGAN?
- A one-shot face stylization model that fine-tunes StyleGAN on a single reference image to apply artistic styles to any face.
- Is JoJoGAN open source?
- Yes — mchong6/JoJoGAN is open source, released under the MIT license.
- What language is JoJoGAN written in?
- mchong6/JoJoGAN is primarily written in Jupyter Notebook.
- How popular is JoJoGAN?
- mchong6/JoJoGAN has 1.4k stars on GitHub.
- Where can I find JoJoGAN?
- mchong6/JoJoGAN is on GitHub at https://github.com/mchong6/JoJoGAN.