Research map / 3D vision / Depth and stereo
PolarAPP: Beyond Polarization Demosaicking for Polarimetric Applications
Yidong Luo, Chenggong Li, Yunfeng Song, Ping Wang, Boxin Shi, Junchao Zhang, Xin Yuan
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 demosaicker (D) in joint training (ablation) | PolarAPP · SfPUEL dataset (SfP) | normal accuracy within 11.25 deg (%): 78.14 -> 92.37 | Improved |
| add task-aware joint training instead of task-agnostic frozen D (ablation) | PolarAPP · SfPUEL dataset (SfP) | normal accuracy within 11.25 deg (%): 76.91 -> 92.37 | Improved |
| add equivalent imaging transformation loss (ablation) | PolarAPP demosaicker · Qiu dataset (AoP) | AoP MAE (deg): 8.011 -> 5.444 | Improved |
| add meta-learned feature alignment over naive joint training (ablation) | PolarAPP · SfPUEL dataset (SfP) | normal accuracy within 11.25 deg (%): 82.44 -> 92.37 | Improved |
| add task refinement stage with frozen demosaicker (ablation) | PolarAPP · SfPUEL dataset (SfP) | mean angular error (deg): 4.68 -> 3.57 | Improved |
| add meta-learned alignment over naive joint training (ablation) | PolarAPP · Qiu dataset (DoLP) | PSNR: 38.43 -> 38.10 | Hurt |
| joint PolarAPP vs PIDSR demosaicker with retrained task network on DfP LPIPS | PolarAPP · PolarRGB (DfP) | LPIPS: 0.124 -> 0.129 | Hurt |
| joint PolarAPP vs retrained PIDSR two-stage on SfP | PolarAPP · SfPUEL dataset (SfP) | mean angular error (deg): 6.78 -> 3.57 | Improved |
| increase alignment weight lambda_fa beyond optimum (100 for SfP, 10 for DfP) (ablation) | PolarAPP · SfP and DfP | task metrics | Hurt |
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.
- SfPUEL — SfPUEL: Shape from Polarization under Unknown Environment Light (NeurIPS 2024) · better on 3 of 3 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
- RGB-to-Polarization Estimation: A New Task and Benchmark Study (NeurIPS 2025 · similarity 0.86)
- PIDSR: Complementary Polarized Image Demosaicing and Super-Resolution (CVPR 2025 · similarity 0.86)
- Polarization Wavefront Lidar: Learning Large Scene Reconstruction from Polarized Wavefronts (CVPR 2024 · similarity 0.86)
- PolarMatte: Fully Computational Ground-Truth-Quality Alpha Matte Extraction for Images and Video using Polarized Screen Matting (CVPR 2024 · similarity 0.86)
- PolarFree: Polarization-based Reflection-Free Imaging (CVPR 2025 · similarity 0.85)
- PolarScale: A Physics-Grounded Benchmark for Radiometrically Consistent RGB-to-Stokes Estimation (NeurIPS 2026 · similarity 0.85)
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.
