Research map / 3D vision / Depth and stereo
S$^3$-NeRF: Neural Reflectance Field from Shading and Shadow under a Single Viewpoint
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 |
|---|---|---|---|
| add shading (BRDF) module instead of RGB MLP (ablation) | S3-NeRF · CHAIR train view | MAE: 32.49 -> 1.83 | Helped |
| add online shadow computation (ablation) | S3-NeRF · CHAIR novel views | MAE: 12.24 -> 5.45 | Helped |
| add surface rendering loss Ls (ablation) | S3-NeRF · CHAIR train view | MAE: 2.48 -> 1.83 | Helped |
| add surface rendering loss Ls (ablation) | S3-NeRF · CHAIR novel views | PSNR: 28.58 -> 26.82 (CHAIR); 28.61 -> 31.53 (BUDDHA) | Mixed |
| ray-marching shadow instead of visibility MLP regression (ablation) | S3-NeRF · CHAIR | Depth L1: 17.14 -> 5.57 | Helped |
| add background shadow modeling (ablation) | S3-NeRF · BUNNY | Depth L1: 34.60 -> 6.82 | Helped |
| add foreground shadow modeling (ablation) | S3-NeRF · BUNNY | Depth L1: 6.75 -> 6.82 | No clear effect |
| increase lights from 4 to 8 (ablation) | S3-NeRF · CHAIR | MAE: 30.39 -> 2.33 | Helped |
| increase lights from 8 to 128 (ablation) | S3-NeRF · CHAIR | MAE: 2.33 -> 1.83 | Helped |
| broaden light range from small to broad (ablation) | S3-NeRF · CHAIR | MAE: 3.92 -> 1.83 | Helped |
| physics-based reflectance field instead of light-conditioned color MLP | NeRF/UNISURF · BUDDHA normals | MAE: 70.12 -> 2.44 | Helped |
Closest papers by meaning
- NeRS: Neural Reflectance Surfaces for Sparse-view 3D Reconstruction in the Wild (NeurIPS 2021 · similarity 0.90)
- Neural Inverse Rendering for General Reflectance Photometric Stereo (ICML 2018 · similarity 0.90)
- UniSDF: Unifying Neural Representations for High-Fidelity 3D Reconstruction of Complex Scenes with Reflections (NeurIPS 2024 · similarity 0.89)
- NeRFPrior: Learning Neural Radiance Field as a Prior for Indoor Scene Reconstruction (CVPR 2025 · similarity 0.89)
- 3D Reconstruction with Generalizable Neural Fields using Scene Priors (ICLR 2024 · similarity 0.89)
- GNeRP: Gaussian-guided Neural Reconstruction of Reflective Objects with Noisy Polarization Priors (ICLR 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.