Research map / 3D vision / Depth and stereo
Towards Accurate Reconstruction of 3D Scene Shape From A Single Monocular Image
Wei Yin, Jianming Zhang, Oliver Wang, Simon Niklaus, Simon Chen, Yifan Liu, Chunhua Shen
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 |
|---|---|---|---|
| recover depth shift with the point cloud module | MiDaS-style relative depth · NYU | AbsRel: 25.8 -> 15.1 | Improved |
| recover depth shift with the point cloud module | relative depth model · KITTI | AbsRel: 23.3 -> 17.5 | Improved |
| add PCM focal length and DPM depth (pinhole model) | Ours · OASIS | LSIV: 1.60 -> 0.52 | Improved |
| replace Hold-Geoffroy focal length with our PCM | MiDaS · OASIS | LSIV: 1.60 -> 1.32 | Improved |
| replace Hold-Geoffroy focal length with our PCM | MiDaS · 2D-3D-S | LSIV: 0.94 -> 0.94 | No clear effect |
| use orthographic camera with our DPM instead of MiDaS | depth model · OASIS | LSIV: 0.63 -> 0.63 | No clear effect |
| add pair-wise normal (PWN) loss (ablation) | Ours (small subset) · iBims-1 | planarity error: 3.91 -> 2.95 | Improved |
| add pair-wise normal (PWN) loss (ablation) | Ours (small subset) · iBims-1 | orientation error: 13.47 -> 11.59 | Improved |
| add pair-wise normal (PWN) loss (ablation) | Ours (small subset) · iBims-1 | DBE acc: 2.05 -> 1.91 | Improved |
| use ILNR instead of MinMax normalization (ablation) | depth loss (small subset) · DIODE | AbsRel: 46.1 -> 34.2 | Improved |
| add Tanh term to ILNR (ablation) | ILNR loss · NYU | AbsRel: 14.3 -> 13.9 | Improved |
| use ILNR instead of SSMAE (ablation) | depth loss (small subset) · KITTI | AbsRel: 18.2 -> 16.1 | Improved |
Closest papers by meaning
- Learning To Recover 3D Scene Shape From a Single Image (CVPR 2021 · similarity 0.96)
- A Simple yet Universal Framework for Depth Completion (NeurIPS 2024 · similarity 0.90)
- Depth2Pose: A Pose-Based Benchmark for Monocular Depth Estimation without Ground-Truth Depth (NeurIPS 2026 · similarity 0.90)
- Depth Pro: Sharp Monocular Metric Depth in Less Than a Second (ICLR 2025 · similarity 0.90)
- RSA: Resolving Scale Ambiguities in Monocular Depth Estimators through Language Descriptions (NeurIPS 2024 · similarity 0.90)
- Seurat: From Moving Points to Depth (CVPR 2025 · similarity 0.90)
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.
