Research map / Image generation / Editing and style transfer
HairNet: Hairstyle Transfer with Pose Changes
Peihao Zhu, Rameen Abdal, John Femiani, Peter Wonka
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 |
|---|---|---|---|
| add HairNet blending network (ablation) | HairNet pipeline · Hairstyle transfer (FFHQ) | FID: 56.4 -> 55.6 | No clear effect |
| add HairNet blending network (ablation) | HairNet pipeline · Hairstyle transfer (FFHQ) | NIQE: 11.82 -> 11.85 | No clear effect |
| add HairNet blending network (ablation) | HairNet pipeline · Hairstyle transfer (FFHQ) | Precision: 95.5% -> 97% | Improved |
| add HairNet blending network (ablation) | HairNet pipeline · Hairstyle transfer (FFHQ) | Recall: 57% -> 60% | Improved |
| add HairNet blending network (ablation) | HairNet pipeline · Hairstyle transfer (FFHQ) | Realism: 1.21 -> 1.26 | Improved |
| add HairNet blending network (ablation) | HairNet pipeline · qualitative figure | visual quality | Improved |
| mix style codes with masked LPIPS optimization (ablation) | HairNet pipeline · qualitative figure | hair and face color preservation | Improved |
| use HairNet instead of Barbershop | hairstyle transfer · User study hair reconstruction | preference rate: 30% -> 70% | Improved |
| use HairNet instead of MichiGAN | hairstyle transfer · User study hair reconstruction | preference rate: 4% -> 96% | Improved |
| use HairNet instead of StyleFusion | hairstyle transfer · User study overall quality | preference rate: 51% -> 49% | No clear effect |
| use HairNet instead of Barbershop | hairstyle transfer · Hairstyle transfer (FFHQ) | FID: 47.34 -> 55.60 | Hurt |
| use HairNet instead of Barbershop | hairstyle transfer · Hairstyle transfer (FFHQ) | Recall: 0.83 -> 0.60 | Hurt |
Closest papers by meaning
- What to Preserve and What to Transfer: Faithful, Identity-Preserving Diffusion-based Hairstyle Transfer (AAAI 2025 · similarity 0.91)
- Stable-Hair: Real-World Hair Transfer via Diffusion Model (AAAI 2025 · similarity 0.91)
- HairFastGAN: Realistic and Robust Hair Transfer with a Fast Encoder-Based Approach (NeurIPS 2024 · similarity 0.91)
- HairDiffusion: Vivid Multi-Colored Hair Editing via Latent Diffusion (NeurIPS 2024 · similarity 0.88)
- Perm: A Parametric Representation for Multi-Style 3D Hair Modeling (ICLR 2025 · similarity 0.88)
- DiffLocks: Generating 3D Hair from a Single Image using Diffusion Models (CVPR 2025 · similarity 0.87)
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.
