Research map / 3D vision / Depth and stereo
Unsupervised Collaborative Learning of Keyframe Detection and Visual Odometry Towards Monocular Deep SLAM
Lu Sheng, Dan Xu, Wanli Ouyang, Xiaogang Wang
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 cycle consistency loss (ablation) | Ours · KITTI Eigen split (80m cap) | abs rel: 0.168 -> 0.139 | Improved |
| add cycle consistency loss (ablation) | Ours · KITTI Eigen split (80m cap) | sq rel: 1.259 -> 1.021 | Improved |
| add cycle consistency loss (ablation) | Ours · KITTI Eigen split (80m cap) | RMSE: 5.937 -> 5.418 | Improved |
| add keyframe selection with online updating over no keyframe selection (ablation) | Ours · KITTI Eigen split (80m cap) | abs rel: 0.181 -> 0.139 | Improved |
| add keyframe selection with online updating over no keyframe selection (ablation) | Ours · KITTI Eigen split (80m cap) | sq rel: 1.587 -> 1.021 | Improved |
| use off-line keyframes pretrained with visual cue only instead of no keyframes (ablation) | Ours · KITTI Eigen split (80m cap) | abs rel: 0.181 -> 0.171 | Improved |
| jointly learn keyframe selector with fixed-length keyframes instead of off-line keyframes (ablation) | Ours · KITTI Eigen split (80m cap) | abs rel: 0.171 -> 0.151 | Improved |
| use online keyframe pool updating instead of fixed-length keyframes (ablation) | Ours · KITTI Eigen split (80m cap) | abs rel: 0.151 -> 0.139 | Improved |
| add keyframe selection with online updating over no keyframe selection (ablation) | Ours · KITTI odometry sequence 09 | absolute trajectory error: 0.018 -> 0.012 | Improved |
| add keyframe selection with online updating over no keyframe selection (ablation) | Ours · KITTI odometry sequence 10 | absolute trajectory error: 0.017 -> 0.010 | Improved |
| use off-line keyframes instead of no keyframes (ablation) | Ours · KITTI odometry sequence 09 | absolute trajectory error: 0.018 -> 0.015 | Improved |
| use fixed-length keyframes instead of off-line keyframes (ablation) | Ours · KITTI odometry sequence 09 | absolute trajectory error: 0.015 -> 0.014 | 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.
- GeoNet — GeoNet: Geometric Neural Network for Joint Depth and Surface Normal Estimation (CVPR 2018) · better on 5 of 6 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
- Unsupervised Learning of Monocular Depth Estimation and Visual Odometry With Deep Feature Reconstruction (CVPR 2018 · similarity 0.89)
- UniSim-SLAM: Feed-Forward SLAM with Unified Sim(3) Optimization (ECCV 2026 · similarity 0.88)
- D3VO: Deep Depth, Deep Pose and Deep Uncertainty for Monocular Visual Odometry (CVPR 2020 · similarity 0.87)
- Learning Monocular Visual Odometry via Self-Supervised Long-Term Modeling (ECCV 2020 · similarity 0.87)
- Deep Virtual Stereo Odometry: Leveraging Deep Depth Prediction for Monocular Direct Sparse Odometry (ECCV 2018 · similarity 0.87)
- Leveraging Consistent Spatio-Temporal Correspondence for Robust Visual Odometry (AAAI 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.
