Research map / 3D vision / Depth and stereo
Mask-ToF: Learning Microlens Masks for Flying Pixel Correction in Time-of-Flight Imaging
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Jointly optimize a spatially varying microlens mask with the refinement network instead of using an open aperture | Mask-ToF · Real ToF scenes | flying pixel count: cuts flying pixel counts in half | Helped |
| Jointly optimize the mask rather than use a hand-designed diameter-5 mask | Mask-ToF · Real ToF scenes | flying pixel count: 30.5% | Helped |
| Use the jointly optimized mask rather than hand-designed masks | Mask-ToF · Synthetic light field test | RMSE: 8.293 -> 7.115; 12.58 -> 7.115; 8.732 -> 7.115 | Helped |
| Add the smooth L1 loss (ablation) | Mask-ToF · Synthetic light field test | RMSE: 7.913 -> 7.115 | Helped |
| Add the Chamfer loss (ablation) | Mask-ToF · Synthetic light field test | Thresh 15mm: 1.194 -> 1.120 | Helped |
| Add the mask as input to the refinement network (ablation) | Mask-ToF · Synthetic light field test | RMSE: 7.353 -> 7.115 | Helped |
| Keep the first hourglass in the refinement network (ablation) | Mask-ToF · Synthetic light field test | RMSE: 7.432 -> 7.115 | Helped |
| Keep the original intermediate channel count rather than doubling it (ablation) | Mask-ToF · Synthetic light field test | RMSE: 7.169 -> 7.115 | Helped |
| Add a global channel using a duplicated, larger-stride first hourglass (ablation) | Mask-ToF · Synthetic light field test | RMSE: 7.115 -> 7.310 | Hurt |
| Use ToFNet instead of the Mask-ToF refinement architecture | Synthetic light field test · Synthetic light field test | RMSE: 7.115 -> 12.19 | Hurt |
| Use the optimized mask rather than a diameter-1 mask | Mask-ToF · Synthetic light field test | MAE: 4.576 -> 1.278 | Helped |
| Use the optimized mask rather than an open-aperture mask | Mask-ToF · Synthetic light field test | Thresh 15mm: 3.558 -> 1.120 | Helped |
Closest papers by meaning
- Deep End-to-End Time-of-Flight Imaging (CVPR 2018 · similarity 0.89)
- Time of the Flight of the Gaussians: Optimizing Depth Indirectly in Dynamic Radiance Fields (CVPR 2025 · similarity 0.87)
- Dynamic Time-Of-Flight (CVPR 2017 · similarity 0.86)
- Snapshot Lidar: Fourier Embedding of Amplitude and Phase for Single-Image Depth Reconstruction (CVPR 2024 · similarity 0.86)
- Learnable Burst-Encodable Time-of-Flight Imaging for High-Fidelity Long-Distance Depth Sensing (NeurIPS 2025 · similarity 0.86)
- Time-Efficient Light-Field Acquisition Using Coded Aperture and Events (CVPR 2024 · 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.