Research map / 3D vision / Depth and stereo
RGB-D SLAM in Dynamic Environments Using Point Correlations
Weichen Dai, Yu Zhang, Li Ping, Zheng Fang, Sebastian Scherer
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 static point determination from point correlations (ablation) | ORB-SLAM2 · TUM RGB-D fr3/walking-static | ATE trans. RMSE (m): 0.4080 -> 0.0108 | Improved |
| add static point determination from point correlations (ablation) | ORB-SLAM2 · TUM RGB-D fr3/walking-halfsphere | ATE trans. RMSE (m): 0.7225 -> 0.0354 | Improved |
| add static point determination from point correlations (ablation) | ORB-SLAM2 · TUM RGB-D fr3/walking-xyz | RPE rot. RMSE (deg/s): 19.7299 -> 2.7413 | Improved |
| add static point determination from point correlations (ablation) | ORB-SLAM2 · TUM RGB-D fr3/sitting-xyz | ATE trans. RMSE (m): 0.0094 -> 0.0091 | Improved |
| add static point determination from point correlations (ablation) | ORB-SLAM2 · TUM RGB-D fr2/desk-person | ATE trans. RMSE (m): 0.0064 -> 0.0075 | Hurt |
| add static point determination from point correlations (ablation) | ORB-SLAM2 · TUM RGB-D fr3/sitting-static | ATE trans. RMSE (m): 0.0077 -> 0.0096 | Hurt |
| add static point determination from point correlations (ablation) | ORB-SLAM2 · TUM RGB-D slightly dynamic sequences | accuracy | Mixed |
| use geometry-only point correlation instead of DynaSLAM (learning-based) | Ours vs DynaSLAM · TUM RGB-D fr3/walking-xyz | ATE trans. RMSE (m): 0.015 -> 0.0874 | Hurt |
| use point-correlation segmentation instead of SPWSLAM | Ours vs SPWSLAM · TUM RGB-D fr3/walking-xyz | ATE trans. RMSE (m): 0.0601 -> 0.0874 | Hurt |
| add static point determination from point correlations (ablation) | ORB-SLAM2 · walking-static | mean time (ms): 31.34 -> 30.65 | Improved |
| use sparse feature graph instead of dense methods | Ours vs DVO · walking-static | mean time (ms): 717.48 -> 30.65 | Improved |
Closest papers by meaning
- Photo-SLAM: Real-time Simultaneous Localization and Photorealistic Mapping for Monocular Stereo and RGB-D Cameras (CVPR 2024 · similarity 0.86)
- Loopy-SLAM: Dense Neural SLAM with Loop Closures (CVPR 2024 · similarity 0.86)
- GS-SLAM: Dense Visual SLAM with 3D Gaussian Splatting (CVPR 2024 · similarity 0.86)
- WildGS-SLAM: Monocular Gaussian Splatting SLAM in Dynamic Environments (CVPR 2025 · similarity 0.86)
- DG-SLAM: Robust Dynamic Gaussian Splatting SLAM with Hybrid Pose Optimization (NeurIPS 2024 · similarity 0.86)
- CTO-SLAM: Contour Tracking for Object-Level Robust 4D SLAM (AAAI 2024 · 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.
