Research map / 3D vision / Depth and stereo
Occlusion-Aware Depth Estimation with Adaptive Normal Constraints
Xiaoxiao Long, Lingjie Liu, Christian Theobalt, Wenping Wang
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 local normal constraint (ablation) | depth network without constraint · 7-Scenes | abs.rel: 0.1925 -> 0.1899 | Mixed |
| add local normal constraint (ablation) | depth network without constraint · 7-Scenes | sq.rel: 0.2350 -> 0.1188 | Improved |
| add global plane normal constraint (ablation) | depth network without constraint · 7-Scenes | abs.rel: 0.1925 -> 0.2096 | Hurt |
| add global plane normal constraint to local constraint (CNM) (ablation) | depth network with local constraint · 7-Scenes | abs.rel: 0.1899 -> 0.1812 | Improved |
| add Combined Normal Map constraint (ablation) | depth network without constraint · 7-Scenes | accuracy sigma<1.25 (%): 71.79 -> 73.17 | Improved |
| add Combined Normal Map constraint (ablation) | depth network without constraint · 7-Scenes normals | normal angle accuracy 11.25 deg (%): 9.877 -> 17.67 | Improved |
| add refinement module (ablation) | CNM depth network · 7-Scenes | abs.rel: 0.1812 -> 0.1726 | Improved |
| add occlusion-aware loss to refinement (ablation) | CNM + refinement · 7-Scenes | abs.rel: 0.1726 -> 0.1669 | Improved |
| add occlusion-aware loss to refinement (ablation) | CNM + refinement · 7-Scenes | accuracy sigma<1.25 (%): 74.77 -> 75.80 | Improved |
| replace virtual normal loss with CNM loss (ablation) | Ours · 7-Scenes normals | mean angle error: 43.08 -> 36.34 | Improved |
| replace virtual normal loss with CNM loss (ablation) | Ours · SUN3D normals | mean angle error: 31.68 -> 29.21 | Improved |
| use middle frame instead of last frame as reference (ablation) | Ours (2 views) · 7-Scenes | abs.rel: 0.1750 -> 0.1669 | Improved |
Closest papers by meaning
- Predicting Sharp and Accurate Occlusion Boundaries in Monocular Depth Estimation Using Displacement Fields (CVPR 2020 · similarity 0.88)
- Normal Assisted Stereo Depth Estimation (CVPR 2020 · similarity 0.88)
- Virtual Normal: Enforcing Geometric Constraints for Accurate and Robust Depth Prediction (TPAMI 2021 · similarity 0.88)
- A Simple yet Universal Framework for Depth Completion (NeurIPS 2024 · similarity 0.87)
- HELVIPAD: A Real-World Dataset for Omnidirectional Stereo Depth Estimation (CVPR 2025 · similarity 0.87)
- Depth2Pose: A Pose-Based Benchmark for Monocular Depth Estimation without Ground-Truth Depth (NeurIPS 2026 · similarity 0.86)
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.
