Research map / 3D vision / Depth and stereo
DEPTHOR: Depth Enhancement from a Practical Light-Weight dToF Sensor and RGB Image
Jijun Xiang, Xuan Zhu, Xianqi Wang, Yu Wang, Hong Zhang, Fei Guo, Xin Yang
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 |
|---|---|---|---|
| train PENet* with the proposed dToF simulation instead of Deltar simulation (ablation) | PENet* · ZJU-L5 | Rel: 0.152 -> 0.093 | Improved |
| train DEPTHOR with the proposed simulation instead of Deltar simulation (ablation) | DEPTHOR · ZJU-L5 | Rel: 0.164 -> 0.075 | Improved |
| include specific-region simulation (SR) (ablation) | dToF simulation · ZJU-L5 | RMSE: 0.566 -> 0.350 | Improved |
| include overall distribution simulation (OD) (ablation) | dToF simulation · ZJU-L5 | RMSE: 0.381 -> 0.350 | Improved |
| include random anomalies simulation (RA) (ablation) | dToF simulation · ZJU-L5 | RMSE: 0.362 -> 0.350 | Improved |
| refine depth with mixed affinity propagation using MDE and UNet features (ablation) | DEPTHOR · Hypersim | RMSE: 0.267 -> 0.248 | Improved |
| refine using MDE features only (ablation) | DEPTHOR · Hypersim | RMSE: 0.267 -> 0.269 | No clear effect |
| add point embedding to affinity propagation (ablation) | DEPTHOR · Hypersim | RMSE: 0.248 -> 0.328 | Hurt |
| use larger relative MDE models (small to large) (ablation) | DEPTHOR · ZJU-L5 | RMSE: 0.350 -> 0.330 | No clear effect |
| use metric-depth MDE (DAv2-MS) instead of relative small (ablation) | DEPTHOR · ZJU-L5 | RMSE: 0.350 -> 0.372 | Hurt |
Closest papers by meaning
- Depth Pro: Sharp Monocular Metric Depth in Less Than a Second (ICLR 2025 · similarity 0.88)
- PatchFusion: An End-to-End Tile-Based Framework for High-Resolution Monocular Metric Depth Estimation (CVPR 2024 · similarity 0.88)
- Depth2Pose: A Pose-Based Benchmark for Monocular Depth Estimation without Ground-Truth Depth (NeurIPS 2026 · similarity 0.87)
- Zero-shot Depth Completion via Test-time Alignment with Affine-invariant Depth Prior (AAAI 2025 · similarity 0.87)
- Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data (CVPR 2024 · similarity 0.87)
- A Simple yet Universal Framework for Depth Completion (NeurIPS 2024 · similarity 0.87)
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.
