mchong6/JoJoGAN

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

1438 stars · 200 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

JoJoGAN is the official PyTorch implementation of a one-shot face stylization method that finetunes a pretrained StyleGAN using GAN inversion from a single reference style image. It transforms portrait photos into styles like anime, art, or sketch while preserving facial details.

Use cases

• stylize a portrait photo into anime style
• apply a single reference image's art style to faces
• convert face photos to sketch style
• experiment with GAN inversion and StyleGAN finetuning
• reproduce one-shot face stylization research
• run face stylization in a notebook or on Hugging Face

When to choose

• you want to restyle face photos using just one style reference image
• you need a research-grade StyleGAN finetuning implementation in PyTorch
• you want anime/art/sketch face stylization with detail preservation

When to avoid

• you need general-purpose image-to-image translation beyond faces
• you need a production service with active maintenance and support
• you lack GPU resources, since finetuning StyleGAN is compute-heavy

Facets

library · maturity maintenance

image-processing machine-learning deep-learning computer-vision image-processing artificial-intelligence deep-learning python gan stylegan face-stylization one-shot-learning anime-style-transfer pytorch research-code gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="mchong6/JoJoGAN")