Research map / 3D vision / Depth and stereo
Learned Off-Aperture Encoding for Wide Field-of-View RGBD Imaging
Haoyu Wei, Xin Liu, Yuhui Liu, Qiang Fu, Wolfgang Heidrich, Edmund Y. M. Lam, Yifan Peng
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 |
|---|---|---|---|
| move DOE off the aperture plane (from on-aperture) with a thin lens (ablation) | App.1 simple lens imaging · Sceneflow (45 deg FoV) | PSNR: 5.35 dB higher PSNR (35.49 dB) | Improved |
| move DOE off the aperture plane with a thin lens (ablation) | App.1 simple lens imaging · Instereo2K (45 deg FoV) | PSNR: 4.41 dB higher PSNR (35.44 dB) | Improved |
| move DOE off the aperture plane with a thin lens (ablation) | App.1 simple lens imaging · Dualpixels (45 deg FoV) | PSNR: 4.13 dB higher PSNR (31.74 dB) | Improved |
| move DOE off the aperture at 60 deg FoV (ablation) | App.1 simple lens imaging · 60 deg FoV (preliminary) | PSNR: 2.5dB improvement | Improved |
| place DOE at intermediate position (about 0.2-0.3 of EFL) rather than at aperture or sensor (ablation) | App.1 simple lens imaging · PSNR vs DOE position (0 to 0.4) | PSNR | Improved |
| vary DOE position between 0.1 and 0.3 of EFL (ablation) | App.1 simple lens imaging · DOE position analysis | PSNR | No clear effect |
| move DOE off-aperture (0.24 EFL) from near-aperture (0.11 EFL) (ablation) | Cooke triplet + DOE · App.2 compound lens RGBD, simulation Sceneflow | PSNR / MAE: 31.00 / 0.037 -> 32.09 / 0.033 | Improved |
| move DOE off-aperture from near-aperture (ablation) | Cooke triplet + DOE · App.2 compound lens RGBD, simulation Dualpixel | PSNR / MAE: 27.95 / 0.025 -> 28.28 / 0.019 | Improved |
| move DOE off-aperture from near-aperture (ablation) | Cooke triplet + DOE · App.2 compound lens RGBD, simulation Instereo2K | PSNR / MAE: 30.51 / 0.032 -> 31.42 / 0.027 | Improved |
| use multi-head decoder instead of U-Net (ablation) | Off-aperture Cooke triplet system · App.2 compound lens RGBD, simulation Sceneflow | PSNR / MAE: 25.27 / 0.046 -> 32.09 / 0.033 | Improved |
| use multi-head decoder instead of U-Net (ablation) | Off-aperture Cooke triplet system · App.2 compound lens RGBD, simulation Dualpixel | PSNR / MAE: 23.69 / 0.027 -> 28.28 / 0.019 | Improved |
| use multi-head decoder instead of U-Net (ablation) | Off-aperture Cooke triplet system · App.2 compound lens RGBD, simulation Instereo2K | PSNR / MAE: 26.06 / 0.124 -> 31.42 / 0.027 | Improved |
Closest papers by meaning
- Learned Binocular-Encoding Optics for RGBD Imaging Using Joint Stereo and Focus Cues (CVPR 2025 · similarity 0.87)
- Passive Snapshot Coded Aperture Dual-Pixel RGB-D Imaging (CVPR 2024 · similarity 0.85)
- A Differentiable Wave Optics Model for End-to-End Computational Imaging System Optimization (ICCV 2025 · similarity 0.85)
- Variable Aperture Light Field Photography: Overcoming the Diffraction-Limited Spatio-Angular Resolution Tradeoff (CVPR 2016 · similarity 0.85)
- Time-Efficient Light-Field Acquisition Using Coded Aperture and Events (CVPR 2024 · similarity 0.83)
- OpticalNet: An Optical Imaging Dataset and Benchmark Beyond the Diffraction Limit (CVPR 2025 · similarity 0.82)
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.
