Depth Estimation on NYU-Depth V2 self-supervised
Metric: Absolute relative error (AbsRel) (lower is better)
Metric: Absolute relative error (AbsRel) (lower is better)
| # | Model↕ | Absolute relative error (AbsRel)▲ | Extra Data | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | IndoorDepth | 0.126 | No | Deeper into Self-Supervised Monocular Indoor Dep... | 2023-12-03 | Code |
| 2 | DistDepth | 0.13 | No | Toward Practical Monocular Indoor Depth Estimation | 2021-12-04 | Code |
| 3 | MonoIndoor | 0.134 | No | MonoIndoor: Towards Good Practice of Self-Superv... | 2021-07-26 | - |
| 4 | StrutDepth | 0.142 | No | StructDepth: Leveraging the structural regularit... | 2021-08-19 | Code |
| 5 | P2Net+PP | 0.147 | No | P$^{2}$Net: Patch-match and Plane-regularization... | 2020-07-15 | Code |
| 6 | Bian et al | 0.157 | No | Unsupervised Scale-consistent Depth Learning fro... | 2021-05-25 | Code |
| 7 | Zhao et al | 0.189 | No | Towards Better Generalization: Joint Depth-Pose ... | 2020-04-03 | Code |
| 8 | Zhou et al | 0.208 | No | Moving Indoor: Unsupervised Video Depth Learning... | 2019-10-20 | - |