Research map / 3D vision / 3D generation
MAIR: Multi-View Attention Inverse Rendering With 3D Spatially-Varying Lighting Estimation
JunYong Choi, SeokYeong Lee, Haesol Park, Seung-Won Jung, Ig-Jae Kim, Junghyun Cho
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 multi-view inverse rendering with MVS depth and confidence | MAIR · OpenRooms FF test | Albedo MSE: 0.569 -> 0.368 | Improved |
| Use multi-view inverse rendering with MVS depth and confidence | MAIR · OpenRooms FF test | Re-rendering MSE: 0.554 -> 0.633 | Hurt |
| Add ContextNet local scene features (ablation) | MAIR material estimation · OpenRooms FF test | Albedo MSE: 0.498 -> 0.423 | Improved |
| Add RefineNet after pixel-wise MVANet (ablation) | MAIR material estimation · OpenRooms FF test | Albedo MSE: 0.661 -> 0.423 | Improved |
| Add the specular radiance feature (ablation) | MAIR material estimation · OpenRooms FF test | Roughness MSE: 3.056 -> 2.739 | Improved |
| Use physically motivated microfacet-based specular encoding (ablation) | MAIR material estimation · OpenRooms FF test | Roughness MSE: 3.085 -> 2.739 | Improved |
| Increase incident-light SG lobe count from 1 to 3 (ablation) | MAIR direct-light estimation · OpenRooms FF | si-MSE: 0.106 -> 0.101 | Improved |
| Increase incident-light SG lobe count from 3 to 6 (ablation) | MAIR direct-light estimation · OpenRooms FF | GPU Memory: 12.72 -> 14.94 GB | Hurt |
| Use depth-projection-based multi-view weights in MVANet (ablation) | MAIR material estimation · Increasing number of views | Material estimation accuracy | Improved |
| Remove multi-view weights (ablation) | MAIR material estimation · Increasing number of views | Material estimation accuracy | No clear effect |
| Provide exitant direct-light volume V_DL instead of global lighting feature (ablation) | MAIR spatially varying lighting estimation · Indirect lighting estimation | Convergence | Improved |
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.
- Lighthouse — Lighthouse: Predicting Lighting Volumes for Spatially-Coherent Illumination (CVPR 2020) · better on 2 of 2 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
- IRIS: Inverse Rendering of Indoor Scenes from Low Dynamic Range Images (CVPR 2025 · similarity 0.90)
- VMINer: Versatile Multi-view Inverse Rendering with Near- and Far-field Light Sources (CVPR 2024 · similarity 0.89)
- RobIR: Robust Inverse Rendering for High-Illumination Scenes (NeurIPS 2024 · similarity 0.86)
- SVG-IR: Spatially-Varying Gaussian Splatting for Inverse Rendering (CVPR 2025 · similarity 0.86)
- Inverse Rendering using Multi-Bounce Path Tracing and Reservoir Sampling (ICLR 2025 · similarity 0.86)
- ESR-NeRF: Emissive Source Reconstruction Using LDR Multi-view Images (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.
