Research map / 3D vision / Depth and stereo
Monocular Depth Estimation with Affinity, Vertical Pooling, and Label Enhancement
Yukang Gan, Xiangyu Xu, Wenxiu Sun, Liang Lin
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 with predicted dense stereo depth as labels instead of image alignment loss (ablation) | Baseline · KITTI Eigen split (no cap) | RMSE: 4.734 -> 4.620 | Improved |
| add residual learning instead of direct depth prediction (ablation) | Baseline‡ · KITTI Eigen split (no cap) | RMSE: 4.620 -> 4.514 | Improved |
| add affinity layer (ablation) | residual model · KITTI Eigen split (no cap) | RMSE: 4.514 -> 4.231 | Improved |
| add affinity layer (ablation) | residual model · KITTI Eigen split (no cap) | ARD: 0.115 -> 0.106 | Improved |
| add squared average pooling (ablation) | R+A model · KITTI Eigen split (no cap) | RMSE: 4.231 -> 4.132 | Improved |
| add horizontal pooling (ablation) | R+A model · KITTI Eigen split (no cap) | RMSE: 4.231 -> 4.141 | Improved |
| use vertical pooling instead of squared pooling (ablation) | R+A model · KITTI Eigen split (no cap) | RMSE: 4.132 -> 4.027 | Improved |
| add fully connected global context (ablation) | R+A+V model · KITTI Eigen split (no cap) | RMSE: 4.027 -> 3.934 | Improved |
| add fully connected global context (ablation) | R+A+V model · KITTI Eigen split (no cap) | delta<1.25^3: 0.982 -> 0.984 | Improved |
| add affinity layer (ablation) | full model · NYU Depth v2 (795 training images) | rms: 0.694 -> 0.631 | Improved |
| use vertical instead of horizontal pooling (ablation) | full model · NYU Depth v2 (795 training images) | rms: 0.652 -> 0.631 | Improved |
| proposed method vs Godard et al. | full model · KITTI Eigen split (cap 0-80 m) | RMSE: 4.935 -> 3.933 | Improved |
Closest papers by meaning
- Learning Monocular Depth Estimation Infusing Traditional Stereo Knowledge (CVPR 2019 · similarity 0.91)
- Self-Supervised Monocular Depth Estimation by Direction-aware Cumulative Convolution Network (ICCV 2023 · similarity 0.89)
- BRNet: Exploring Comprehensive Features for Monocular Depth Estimation (ECCV 2022 · similarity 0.89)
- Distilling Monocular Foundation Model for Fine-grained Depth Completion (CVPR 2025 · similarity 0.89)
- Mind The Edge: Refining Depth Edges in Sparsely-Supervised Monocular Depth Estimation (CVPR 2024 · similarity 0.89)
- Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data (CVPR 2024 · similarity 0.88)
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.
