Research map / 3D vision / Face and shape
PRM: Photometric Stereo based Large Reconstruction Model
Wenhang Ge, Jiantao Lin, Guibao Shen, Jiawei Feng, Tao Hu, Xinli Xu, Ying-Cong 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 |
|---|---|---|---|
| use PRM instead of InstantMesh | InstantMesh · GSO | Chamfer Distance: 0.076 -> 0.050 | Improved |
| use PRM instead of InstantMesh | InstantMesh · GSO | PSNR: 19.988 -> 25.125 | Improved |
| use PRM instead of InstantMesh | InstantMesh · GSO | FS@0.1: 0.931 -> 0.981 | Improved |
| add PBR with differentiable rendering instead of direct shading color MLP (ablation) | PRM · GSO (ablation) | Chamfer Distance: 0.089 -> 0.066 | Improved |
| add lighting map supervision (ablation) | PRM · GSO (ablation) | Chamfer Distance: 0.073 -> 0.066 | Improved |
| add shading color and light map supervision to albedo (ablation) | PRM · GSO (ablation) | Chamfer Distance: 0.099 -> 0.066 | Improved |
| add material variation during training (ablation) | PRM · GSO (ablation) | Chamfer Distance: 0.089 -> 0.066 | Improved |
| add material variation during training (ablation) | PRM · GSO (ablation) | PSNR: 19.662 -> 20.992 | Improved |
| add more input camera views (1 to 8) (ablation) | PRM · GSO | Chamfer Distance and F-Score | Improved |
| randomize materials and lighting of input images | InstantMesh · GSO (GT multi-view input) | Chamfer Distance: 0.048 -> 0.061 | Hurt |
| randomize materials and lighting of input images | PRM · GSO (GT multi-view input) | Chamfer Distance: 0.043 -> 0.053 | Hurt |
| randomize materials and lighting of input images | PRM · GSO (GT multi-view input) | PSNR: 26.377 -> 24.602 | Hurt |
Closest papers by meaning
- Fast Light-Weight Near-Field Photometric Stereo (CVPR 2022 · similarity 0.88)
- MERLiN: Single-Shot Material Estimation and Relighting for Photometric Stereo (ECCV 2024 · similarity 0.88)
- PX-NET: Simple and Efficient Pixel-Wise Training of Photometric Stereo Networks (ICCV 2021 · similarity 0.87)
- Self-Calibrating Photometric Stereo by Neural Inverse Rendering (ECCV 2022 · similarity 0.87)
- Photometric Mesh Optimization for Video-Aligned 3D Object Reconstruction (CVPR 2019 · similarity 0.87)
- RNb-NeuS: Reflectance and Normal-based Multi-View 3D Reconstruction (CVPR 2024 · similarity 0.86)
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.
