Research map / 3D vision / Depth and stereo
High-Precision Self-Supervised Monocular Depth Estimation with Rich-Resource Prior
Jianbing Shen, Wencheng Han
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 Prior Depth Fusion with rich-resource reference features (ablation) | DIFFNet baseline · KITTI Eigen split (640x192, M) | RMSE: 4.483 -> 4.284 | Improved |
| add attention guided feature selection to shrink reference set (ablation) | RPrDepth · KITTI Eigen split (640x192, M) | RMSE: 4.284 -> 4.240 | Improved |
| add Rich-resource Guided Loss (ablation) | DIFFNet baseline · KITTI Eigen split (640x192, M) | Abs Rel: 0.102 -> 0.100 | Improved |
| combine all three components (ablation) | DIFFNet baseline · KITTI Eigen split (640x192, M) | Abs Rel: 0.102 -> 0.097 | Improved |
| reduce reference features to about 1% via feature selection (ablation) | RPrDepth · inference reference set | reference pixels: 2,560,000 -> 25,000 pixels | Improved |
| single low-resolution image model with rich-resource prior versus multi-frame ManyDepth | RPrDepth · KITTI Eigen split (640x192, M) | RMSE: 4.459 -> 4.279 | Improved |
| RPrDepth versus ManyDepth on Cityscapes RMSE | RPrDepth · Cityscapes | RMSE: 6.223 -> 6.243 | Hurt |
| apply method to unseen-domain Make3D without retraining | RPrDepth · Make3D | Abs Rel: 0.302 -> 0.288 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- BRNet — BRNet: Exploring Comprehensive Features for Monocular Depth Estimation (ECCV 2022) · better on 3 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Digging Into Self-Supervised Monocular Depth Estimation (ICCV 2019 · similarity 0.91)
- Excavating the Potential Capacity of Self-Supervised Monocular Depth Estimation (ICCV 2021 · similarity 0.91)
- HR-Depth: High Resolution Self-Supervised Monocular Depth Estimation (AAAI 2021 · similarity 0.91)
- Self-Supervised Monocular Trained Depth Estimation Using Self-Attention and Discrete Disparity Volume (CVPR 2020 · similarity 0.90)
- The Temporal Opportunist: Self-Supervised Multi-Frame Monocular Depth (CVPR 2021 · similarity 0.90)
- Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data (CVPR 2024 · similarity 0.89)
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.
