Research map / 3D vision / NeRF and radiance fields
PS-NeRF: Neural Inverse Rendering for Multi-View Photometric Stereo
Wenqi Yang, Guanying Chen, Chaofeng Chen, Zhenfang Chen, Kwan-Yee K. Wong
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 multi-light normal regularization (ablation) | NeRF · DiLiGenT-MV READING | Normal MAE: 55.55 -> 14.01 | Improved |
| add multi-light normal regularization (ablation) | UNISURF · DiLiGenT-MV READING | Normal MAE: 19.68 -> 9.58 | Improved |
| add visibility modeling (ablation) | PS-NeRF Stage II · Synth-PS BUNNY | PSNR: 21.29 -> 25.88 | Improved |
| optimize light directions instead of fixing them (ablation) | PS-NeRF Stage II · Synth-PS BUNNY | PSNR: 24.04 -> 25.88 | Improved |
| add joint optimization (Stage II) (ablation) | PS-NeRF · DiLiGenT-MV READING | Normal MAE: 9.58 -> 8.87 | Improved |
| refine lights by joint optimization | PS-NeRF · DiLiGenT-MV average | Light direction MAE: 6.89 -> 2.95 | Improved |
| use specular SG basis instead of Microfacet roughness (ablation) | PS-NeRF · DiLiGenT-MV + BUNNY average | Normal MAE: 6.49 -> 6.24 | Improved |
| increase training views from 5 to 30 | UNISURF · Synth-Env BUNNY | Normal MAE: 20.31 -> 9.11 | Improved |
| increase training views from 5 to 15 | PS-NeRF · Synth-Env BUNNY | Normal MAE: 6.96 -> 5.17 | Improved |
| increase lights from 2 to 8 | PS-NeRF · DiLiGenT-MV READING (5 views) | Normal MAE: 24.48 -> 11.52 | Improved |
| increase lights from 4 to 96 | PS-NeRF · DiLiGenT-MV | Normal MAE: 6.75 vs. 5.18 | No clear effect |
| increase views from 10 to 15 (8 lights) | PS-NeRF · DiLiGenT-MV READING | Normal MAE: 9.63 -> 9.99 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- UNISURF — UNISURF: Unifying Neural Implicit Surfaces and Radiance Fields for Multi-View Reconstruction (ICCV 2021) · better on 5 of 5 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
- RNb-NeuS: Reflectance and Normal-based Multi-View 3D Reconstruction (CVPR 2024 · similarity 0.90)
- MVCPS-NeuS: Multi-view Constrained Photometric Stereo for Neural Surface Reconstruction (CVPR 2024 · similarity 0.90)
- Sparse Views Near Light: A Practical Paradigm for Uncalibrated Point-light Photometric Stereo (CVPR 2024 · similarity 0.89)
- A Pre-convolved Representation for Plug-and-Play Neural Illumination Fields (AAAI 2024 · similarity 0.88)
- NeRFPrior: Learning Neural Radiance Field as a Prior for Indoor Scene Reconstruction (CVPR 2025 · similarity 0.88)
- Inverse Rendering of Glossy Objects via the Neural Plenoptic Function and Radiance Fields (CVPR 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.
