Research map / 3D vision / Depth and stereo
Moving Indoor: Unsupervised Video Depth Learning in Challenging Environments
Junsheng Zhou, Yuwang Wang, Kaihuai Qin, Wenjun Zeng
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 sparse-seed propagation to SF-Net (ablation) | SF-Net · NYU validation | EPE: 7.409 -> 3.602 | Improved |
| add propagation in SF-Net flow supervision (ablation) | DepthNet + F-PoseNet · NYU Depth V2 test | delta<1.25: 0.596 -> 0.674 | Improved |
| feed flow instead of RGB to PoseNet (ablation) | DepthNet with propagated flow · NYU Depth V2 test | delta<1.25: 0.578 -> 0.674 | Improved |
| add pure-rotation filtering (ablation) | DepthNet + F-PoseNet (flow w/o propagation) · NYU Depth V2 test | delta<1.25: 0.511 -> 0.596 | Improved |
| add flow supervision with pure-rotation filtering (ablation) | RGB-based baseline · NYU Depth V2 test | delta<1.25: 0.511 -> 0.596 | Improved |
| add flow supervision and F-PoseNet (ablation) | baseline · Scannet subset | delta<1.25: 0.631 -> 0.710 | Improved |
| use flow input to PoseNet on top of flow supervision (ablation) | Scannet model · Scannet subset | delta<1.25: 0.682 -> 0.710 | Improved |
| add flow supervision (ablation) | baseline · KITTI | delta<1.25: 0.833 -> 0.837 | No clear effect |
| add F-PoseNet (ablation) | baseline · KITTI | rel: 0.132 -> 0.130 | Improved |
| use PWC-Net instead of SF-Net (ablation) | flow estimator · NYU validation | EPE: 3.602 -> 3.279 | Mixed |
Closest papers by meaning
- Depth Prediction without the Sensors: Leveraging Structure for Unsupervised Learning from Monocular Videos (AAAI 2019 · similarity 0.89)
- Depth From Videos in the Wild: Unsupervised Monocular Depth Learning From Unknown Cameras (ICCV 2019 · similarity 0.89)
- Competitive Collaboration: Joint Unsupervised Learning of Depth, Camera Motion, Optical Flow and Motion Segmentation (CVPR 2019 · similarity 0.88)
- P²Net: Patch-match and Plane-regularization for Unsupervised Indoor Depth Estimation (ECCV 2020 · similarity 0.88)
- Auto-Rectify Network for Unsupervised Indoor Depth Estimation (TPAMI 2021 · similarity 0.88)
- Seurat: From Moving Points to Depth (CVPR 2025 · similarity 0.85)
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.
