Research map / 3D vision / Face and shape
Deep Inverse Shading: Consistent Albedo and Surface Detail Recovery via Generative Refinement
Jiacheng Wu, Ruiqi Zhang, Jie Chen
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 normal conversion module O2N (ablation) | DIS · ZJUMoCap and THUman2.0 | normal degree: 24.47 -> 15.34 | Improved |
| add O2N instead of splitting the SMPL mesh (6,890 to 27,554 vertices) (ablation) | DIS · ZJUMoCap and THUman2.0 | normal degree: 20.36 -> 15.34 | Improved |
| use O2N instead of Mesh Split for training cost (ablation) | DIS · ZJUMoCap and THUman2.0 | training time per epoch: 20+ -> 5 Min./Epoch | Improved |
| add the de-shading module (ablation) | DIS · Surface normal | normal degree: 16.72 -> 15.89 | Improved |
| add O2N to PBR (ablation) | DIS · SyntheticHuman | visibility map detail | Improved |
| add de-shading module (ablation) | DIS · synthetic and real humans | albedo | Improved |
| DIS vs RA relighting | DIS · SyntheticHuman++ | relighting PSNR: 34.59 -> 35.32 | Improved |
| DIS vs RA training time | DIS · SyntheticHuman++ | training time: 50+ hours -> 3 hours | Improved |
| DIS vs RA memory | DIS · rendering efficiency | memory (GB): 10.89 -> 3.8 | Improved |
| DIS vs 3DGS-Avatar rendering speed | DIS · rendering efficiency | speed (FPS): 50+ -> 12.16 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- Relighting4D — Relighting4D: Neural Relightable Human from Videos (ECCV 2022) · better on 9 of 9 shared results
- 3DGS-Avatar — 3DGS-Avatar: Animatable Avatars via Deformable 3D Gaussian Splatting (CVPR 2024) · better on 1 of 1 shared result
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
- IntrinsicAvatar: Physically Based Inverse Rendering of Dynamic Humans from Monocular Videos via Explicit Ray Tracing (CVPR 2024 · similarity 0.90)
- Synthetic Prior for Few-Shot Drivable Head Avatar Inversion (CVPR 2025 · similarity 0.89)
- En3D: An Enhanced Generative Model for Sculpting 3D Humans from 2D Synthetic Data (CVPR 2024 · similarity 0.89)
- Human-3Diffusion: Realistic Avatar Creation via Explicit 3D Consistent Diffusion Models (NeurIPS 2024 · similarity 0.89)
- DiffHuman: Probabilistic Photorealistic 3D Reconstruction of Humans (CVPR 2024 · similarity 0.88)
- Vid2Avatar-Pro: Authentic Avatar from Videos in the Wild via Universal Prior (CVPR 2025 · 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.
