Research map / 3D vision / NeRF and radiance fields
Learning Indoor Inverse Rendering With 3D Spatially-Varying Lighting
Zian Wang, Jonah Philion, Sanja Fidler, Jan Kautz
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 Joint Re-prediction module (ablation) | Ours · InteriorNet | albedo si-MSE: 0.0190 -> 0.0175 | Improved |
| add Joint Re-prediction module (ablation) | Ours · InteriorNet | normal angular error: 19.09 -> 18.40 | Improved |
| add Joint Re-prediction module (ablation) | Ours · InteriorNet | depth si-MSE: 0.217 -> 0.181 | Improved |
| add lighting-related properties to Joint Re-prediction (ablation) | Ours · InteriorNet | depth si-MSE: 0.189 -> 0.181 | Improved |
| add Joint Re-prediction module (ablation) | Ours · NYUv2 | depth si-MSE: 0.3196 -> 0.2827 | Improved |
| add Joint Re-prediction module (ablation) | Ours · IIW | WHDR: 18.7 -> 18.2 | Improved |
| add visible-view loss (ablation) | Ours · InteriorNet lighting | PSNR: 16.43 -> 17.06 | Improved |
| add alpha regularization loss (ablation) | Ours · InteriorNet lighting | PSNR: 17.06 -> 17.33 | Improved |
| add re-rendering loss to lighting training (ablation) | Ours · InteriorNet lighting | PSNR: 17.33 -> 17.37 | No clear effect |
| add re-rendering loss (ablation) | Ours · InteriorNet | re-rendering MSE: 2.18 -> 0.89 | Improved |
| use Volumetric Spherical Gaussian instead of RGB-alpha volume (ablation) | Ours · InteriorNet lighting | PSNR: 16.94 -> 17.37 | Improved |
| use Volumetric Spherical Gaussian instead of RGB-alpha volume (ablation) | Ours · NYUv2 | re-rendering MSE: 2.72 -> 2.33 | Improved |
Closest papers by meaning
- Inverse Rendering for Complex Indoor Scenes: Shape, Spatially-Varying Lighting and SVBRDF From a Single Image (CVPR 2020 · similarity 0.94)
- Physically-Based Editing of Indoor Scene Lighting from a Single Image (ECCV 2022 · similarity 0.92)
- Multi-View Inverse Rendering for Large-Scale Real-World Indoor Scenes (CVPR 2023 · similarity 0.91)
- Neural Inverse Rendering of an Indoor Scene From a Single Image (ICCV 2019 · similarity 0.91)
- PhyIR: Physics-Based Inverse Rendering for Panoramic Indoor Images (CVPR 2022 · similarity 0.90)
- IRIS: Inverse Rendering of Indoor Scenes from Low Dynamic Range Images (CVPR 2025 · similarity 0.89)
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.
