Research map / Point clouds / Point cloud classification
Point-SLAM: Dense Neural Point Cloud-based SLAM
Erik Sandström, Yue Li, Luc Van Gool, Martin R. Oswald
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 non-linear preprocessing network F_theta on color features (ablation) | Point-SLAM · Replica Room 0 | PSNR (dB): 27.41 -> 32.09 | Improved |
| add F_theta when tracking errors are high | Point-SLAM · TUM-RGBD, ScanNet | rendering quality | No clear effect |
| add RGB to tracking (ablation) | Point-SLAM · Replica Room 0 | ATE RMSE (cm): 0.67 -> 0.36 | Improved |
| add RGB to mapping (ablation) | Point-SLAM · Replica Room 0 | F1 (%): 91.37 -> 91.49 | No clear effect |
| increase upper search radius r_u (more compression) | Point-SLAM · Replica Room 0 | ATE RMSE, depth L1, PSNR | Mixed |
| use adaptive neural point cloud instead of NICE-SLAM voxel grid | dense neural RGBD SLAM · Replica rendering | PSNR (dB): 24.42 -> 35.17 | Improved |
| use Point-SLAM on ScanNet instead of NICE-SLAM | Point-SLAM · ScanNet tracking | ATE RMSE (cm): 10.70 -> 12.19 | Hurt |
| use Point-SLAM instead of traditional sparse trackers | TUM-RGBD tracking · TUM-RGBD | ATE RMSE (cm): ORB-SLAM2 1.98 vs 8.92 | Hurt |
Closest papers by meaning
- Loopy-SLAM: Dense Neural SLAM with Loop Closures (CVPR 2024 · similarity 0.92)
- Benchmarking Implicit Neural Representation and Geometric Rendering in Real-Time RGB-D SLAM (CVPR 2024 · similarity 0.91)
- IBD-SLAM: Learning Image-Based Depth Fusion for Generalizable SLAM (CVPR 2024 · similarity 0.89)
- DeepPointMap: Advancing LiDAR SLAM with Unified Neural Descriptors (AAAI 2024 · similarity 0.88)
- SLAM3R: Real-Time Dense Scene Reconstruction from Monocular RGB Videos (CVPR 2025 · similarity 0.87)
- Implicit Event-RGBD Neural SLAM (CVPR 2024 · similarity 0.87)
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.
