Research map / 3D vision / Depth and stereo
StructDepth: Leveraging the Structural Regularities for Self-Supervised Indoor Depth Estimation
Boying Li, Yuan Huang, Zeyu Liu, Danping Zou, Wenxian Yu
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 Manhattan normal loss (ablation) | P2Net-finetune base · NYUv2 | AbsRel: 0.147 -> 0.143 | Improved |
| add co-planar loss (ablation) | P2Net-finetune base · NYUv2 | AbsRel: 0.147 -> 0.144 | Improved |
| add co-planar loss on top of Manhattan normal loss (ablation) | StructDepth · NYUv2 | AbsRel: 0.143 -> 0.142 | Improved |
| add Manhattan normal loss on top of co-planar loss (ablation) | StructDepth · NYUv2 | AbsRel: 0.144 -> 0.142 | Improved |
| add structural losses (ablation) | Monodepth2 architecture · NYUv2 | AbsRel: 0.159 -> 0.151 | Improved |
| add structural losses (ablation) | P2Net architecture · NYUv2 | AbsRel: 0.147 -> 0.142 | Improved |
| add structural losses (ablation) | Monodepth2 architecture · KITTI | AbsRel: 0.115 -> 0.120 | Hurt |
| add structural losses (ablation) | P2Net architecture · KITTI | AbsRel: 0.121 -> 0.120 | No clear effect |
| add post processing | StructDepth · NYUv2 | AbsRel: 0.142 -> 0.140 | Improved |
| use StructDepth instead of P2Net | self-supervised depth · NYUv2 | AbsRel: 0.150 -> 0.142 | Improved |
| use StructDepth instead of P2Net | self-supervised depth · ScanNet | AbsRel: 0.175 -> 0.165 | Improved |
| use StructDepth instead of P2Net | self-supervised depth · InteriorNet | AbsRel: 0.346 -> 0.330 | Improved |
Closest papers by meaning
- Toward Practical Monocular Indoor Depth Estimation (CVPR 2022 · similarity 0.90)
- GasMono: Geometry-Aided Self-Supervised Monocular Depth Estimation for Indoor Scenes (ICCV 2023 · similarity 0.90)
- StructNeRF: Neural Radiance Fields for Indoor Scenes With Structural Hints (TPAMI 2023 · similarity 0.89)
- Geometric Structure Based and Regularized Depth Estimation From 360 Indoor Imagery (CVPR 2020 · similarity 0.88)
- GeoDepth: From Point-to-Depth to Plane-to-Depth Modeling for Self-Supervised Monocular Depth Estimation (CVPR 2025 · similarity 0.87)
- Shading Meets Motion: Self-supervised Indoor 3D Reconstruction Via Simultaneous Shape-from-Shading and Structure-from-Motion (CVPR 2025 · 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.
