Research map / 3D vision / Depth and stereo
Neural Multi-View Self-Calibrated Photometric Stereo without Photometric Stereo Cues
Xu Cao, Takafumi Taketomi
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 angular encoding of normal-view-light directions (ablation) | Weighted-L1 pipeline · DiLiGenT-MV Pot2 32L18V ablation | Light direction MAE (deg): 16.79 -> 2.67 | Improved |
| add angular encoding (ablation) | Weighted-L1 pipeline · DiLiGenT-MV Pot2 32L18V ablation | Mesh Chamfer distance (mm): 0.24 -> 0.16 | Improved |
| add angular encoding (ablation) | Weighted-L1 pipeline · DiLiGenT-MV Pot2 32L18V ablation | Normal MAE (deg): 5.66 -> 4.86 | Improved |
| add angular encoding (ablation) | Weighted-L1 pipeline · DiLiGenT-MV Pot2 32L18V ablation | Test PSNR: 24.98 -> 32.40 | Improved |
| use weighted L1 instead of L2 (ablation) | Pipeline with AE · DiLiGenT-MV Pot2 32L18V ablation | Normal MAE (deg): 11.45 -> 4.86 | Improved |
| use weighted L1 instead of L2 (ablation) | Pipeline with AE · DiLiGenT-MV Pot2 32L18V ablation | Test PSNR: 25.31 -> 32.40 | Improved |
| use weighted L1 instead of weighted L2 (ablation) | Pipeline with AE · DiLiGenT-MV Pot2 32L18V ablation | Mesh Chamfer distance (mm): 0.67 -> 0.16 | Improved |
| use weighted L1 instead of weighted L2 (ablation) | Pipeline with AE · DiLiGenT-MV Pot2 32L18V ablation | Normal MAE (deg): 24.97 -> 4.86 | Improved |
| add weighting to L1 color loss (ablation) | Pipeline with AE · DiLiGenT-MV Pot2 32L18V ablation | Mesh Chamfer distance (mm): 0.27 -> 0.16 | Improved |
| add weighting to L1 color loss (ablation) | Pipeline with AE · DiLiGenT-MV Pot2 32L18V ablation | Light direction MAE (deg): 5.34 -> 2.67 | Improved |
| end-to-end optimization vs PS-NeRF initialized lights | Ours vs PS-NeRF · DiLiGenT-MV Reading (32 lights) | Light direction MAE (deg): 6.21 -> 1.22 | Improved |
| reduce lights from 96 to 4 | Ours · DiLiGenT-MV Reading | Light direction MAE (deg): 2.09 -> 1.16 | No clear effect |
Closest papers by meaning
- Self-Calibrating Photometric Stereo by Neural Inverse Rendering (ECCV 2022 · similarity 0.93)
- PS-NeRF: Neural Inverse Rendering for Multi-View Photometric Stereo (ECCV 2022 · similarity 0.93)
- Uncalibrated Neural Inverse Rendering for Photometric Stereo of General Surfaces (CVPR 2021 · similarity 0.92)
- DANI-Net: Uncalibrated Photometric Stereo by Differentiable Shadow Handling, Anisotropic Reflectance Modeling, and Neural Inverse Rendering (CVPR 2023 · similarity 0.91)
- Sparse Views Near Light: A Practical Paradigm for Uncalibrated Point-light Photometric Stereo (CVPR 2024 · similarity 0.91)
- RNb-NeuS: Reflectance and Normal-based Multi-View 3D Reconstruction (CVPR 2024 · similarity 0.90)
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.
