Research map / 3D vision / Depth and stereo
Structure-Aware Consistency Priors for Shape from Polarization in Complex Media
Kaimin Yu, Puyun Wang, Huayang He, Xianyu Wu
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 the structure-aware polarization consistency prior (ablation) | IceSfP · IceSfP dataset | MAE (degrees): 17.84 -> 16.01 | Improved |
| add the cross-modal reliability attention (CRA) module (ablation) | IceSfP · IceSfP dataset | MAE (degrees): 17.44 -> 16.01 | Improved |
| add the SPADE module in the decoder (ablation) | IceSfP · IceSfP dataset | MAE (degrees): 17.45 -> 16.01 | Improved |
| add the consistency prior (ablation) | IceSfP · IceSfP dataset | accuracy within 11.25 degrees: 34.84% -> 41.92% | Improved |
| set consistency prior weight alpha to 0.5 versus other values 0.1-0.9 (ablation) | consistency prior · IceSfP dataset | MAE (degrees) | Improved |
| use IceSfP instead of strongest baseline | IceSfP · IceSfP dataset | MAE (degrees): 18.75 -> 16.01 | Improved |
| use IceSfP instead of strongest baseline on external datasets | IceSfP · DeepSfP dataset and TransSfP dataset | MAE (degrees): 17.50 -> 16.33 (DeepSfP data); 16.05 -> 15.59 (TransSfP data) | Improved |
| use IceSfP on TransSfP dataset, accuracy within 22.5 and 30 degrees | IceSfP · TransSfP dataset | accuracy within 22.5 degrees: 79.37% vs 80.31% (Attention U2Net) | Mixed |
Closest papers by meaning
- Shape From Polarization for Complex Scenes in the Wild (CVPR 2022 · similarity 0.90)
- Shape from Polarization with Distant Lighting Estimation (TPAMI 2023 · similarity 0.90)
- SfPUEL: Shape from Polarization under Unknown Environment Light (NeurIPS 2024 · similarity 0.87)
- Learning Accurate 3D Shape Based on Stereo Polarimetric Imaging (CVPR 2023 · similarity 0.87)
- PolarScale: A Physics-Grounded Benchmark for Radiometrically Consistent RGB-to-Stokes Estimation (NeurIPS 2026 · similarity 0.86)
- RGB-to-Polarization Estimation: A New Task and Benchmark Study (NeurIPS 2025 · similarity 0.84)
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.
