Research map / Image generation / Editing and style transfer
JoJoGAN: One Shot Face Stylization
Min Jin Chong, David Forsyth
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use discriminator-activation perceptual loss instead of LPIPS at 256 or 1024 resolution (ablation) | JoJoGAN · One-shot face stylization (qualitative) | detail preservation | Improved |
| add ArcFace identity loss (ablation) | JoJoGAN · References that distort identity | identity preservation | Improved |
| use mean style code instead of GAN inversion for extreme (animal) references (ablation) | JoJoGAN · Extreme style references | stylization quality | Improved |
| use multiple style references (multi-shot) instead of one (ablation) | JoJoGAN · Face stylization (qualitative) | stylization quality | Improved |
| use ReStyle inversion instead of II2S (ablation) | JoJoGAN · Inversion choice (qualitative) | input identity preservation | Improved |
| use JoJoGAN instead of non-DST methods (user study) | one-shot face stylization · User study 1 | preference rate: 80.6% | Improved |
| use JoJoGAN instead of DST (user study) | one-shot face stylization · User study 2 | preference rate: 74% | Improved |
| too small a set of mixed style codes S (ablation) | JoJoGAN · Failure analysis | output quality | Hurt |
Later papers that compared with this method
1 later accepted paper reports JoJoGAN as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Deformable One-shot Face Stylization via DINO Semantic Guidance (CVPR 2024 · better on 2 of 3 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Face Generation and Editing With StyleGAN: A Survey (TPAMI 2024 · similarity 0.88)
- BlendGAN: Implicitly GAN Blending for Arbitrary Stylized Face Generation (NeurIPS 2021 · similarity 0.86)
- StyleGANEX: StyleGAN-Based Manipulation Beyond Cropped Aligned Faces (ICCV 2023 · similarity 0.86)
- StyleGAN2 Distillation for Feed-forward Image Manipulation (ECCV 2020 · similarity 0.85)
- Fine-Grained Face Swapping via Regional GAN Inversion (CVPR 2023 · similarity 0.85)
- The Devil is in the Details: StyleFeatureEditor for Detail-Rich StyleGAN Inversion and High Quality Image Editing (CVPR 2024 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
