Research map / 3D vision / Depth and stereo
HairStep: Transfer Synthetic to Real Using Strand and Depth Maps for Single-View 3D Hair Modeling
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Replace the undirected orientation map with HairStep | NeuralHDHair* · USC-HairSalon | orientation error: 0.1324 -> 0.0658 | Helped |
| Replace the undirected orientation map with HairStep | NeuralHDHair* · USC-HairSalon | occupancy accuracy: 82.59\% -> 86.77\% | Helped |
| Replace the undirected orientation map with HairStep | NeuralHDHair* · HiSa and HiDa | HairRida: 70.67\% -> 76.79\% | Helped |
| Replace the undirected orientation map with HairStep | NeuralHDHair* · HiSa and HiDa | HairSale: 19.6 -> 16.36 | Helped |
| Replace the undirected orientation map with HairStep | DynamicHair · USC-HairSalon | orientation error: 0.1352 -> 0.1174 | Helped |
| Replace the undirected orientation map with HairStep | HairNet · USC-HairSalon | orientation error: 0.02349 -> 0.02184 | Helped |
| Replace the undirected orientation map with HairStep | DynamicHair · HiSa and HiDa | HairSale: 32.66 -> 27.51 | Helped |
| Replace the undirected orientation map with HairStep | HairNet · HiSa and HiDa | HairSale: 31.97 -> 27.68 | Helped |
| Predict directed strand maps instead of using Gabor filters | Strand map extraction · HiSa test | HairSale: 18.4 -> 12.3 | Helped |
| Use domain-adaptive depth estimation instead of pseudo depth | Depth estimation · HiSa and HiDa | HairRida: 80.47\% -> 85.20\% | Helped |
| Add synthetic-domain pseudo-depth supervision to weak-label depth estimation (ablation) | HairStep depth representation · HiSa and HiDa | HairRida: 75.8\% -> 76.79\% | Helped |
| Use domain-adaptive depth instead of pseudo depth (ablation) | HairStep depth representation · HiSa and HiDa | HairRida: 73.57\% -> 76.79\% | Helped |
Closest papers by meaning
- DiffLocks: Generating 3D Hair from a Single Image using Diffusion Models (CVPR 2025 · similarity 0.92)
- MonoHair: High-Fidelity Hair Modeling from a Monocular Video (CVPR 2024 · similarity 0.91)
- Perm: A Parametric Representation for Multi-Style 3D Hair Modeling (ICLR 2025 · similarity 0.91)
- Dr.Hair: Reconstructing Scalp-Connected Hair Strands without Pre-Training via Differentiable Rendering of Line Segments (CVPR 2024 · similarity 0.90)
- Text-Conditioned Generative Model of 3D Strand-based Human Hairstyles (CVPR 2024 · similarity 0.89)
- Doubly Hierarchical Geometric Representations for Strand-based Human Hairstyle Generation (NeurIPS 2024 · similarity 0.88)
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.