Research map / Point clouds / Point cloud classification
SLIM: Self-Supervised LiDAR Scene Flow and Motion Segmentation
Stefan Andreas Baur, David Josef Emmerichs, Frank Moosmann, Peter Pinggera, Björn Ommer, Andreas Geiger
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 classification logits and ego-motion aggregation (self-supervised) (ablation) | RAFT + PointPillars · nuScenes ablation (AEE 50-50) self-supervised | AEE 50-50: 0.1551 -> 0.0987 | Improved |
| add classification logits and ego-motion aggregation (supervised) (ablation) | RAFT + PointPillars · nuScenes ablation (AEE 50-50) supervised | AEE 50-50: 0.1401 -> 0.0704 | Improved |
| add confidence weights to the aggregation (self-supervised) (ablation) | SLIM · nuScenes ablation (AEE 50-50) self-supervised | AEE 50-50: 0.0987 -> 0.1256 | Hurt |
| add confidence weights to the aggregation (supervised) (ablation) | SLIM · nuScenes ablation (AEE 50-50) supervised | AEE 50-50: 0.0704 -> 0.0600 | Improved |
| evaluate on full point cloud (all points) instead of 8192 points (ablation) | SLIM · KITTI-SF | AEE: 0.1207 -> 0.0668 | Improved |
| evaluate on full point cloud instead of 8192 points (ablation) | PointPWCNet · KITTI-SF | AEE: 0.3466 -> 0.3648 | Hurt |
| evaluate on full point cloud instead of 8192 points (ablation) | PoseFlowNet · KITTI-SF | AEE: 0.3009 -> 0.3256 | Hurt |
| train self-supervised on KITTI-RL instead of nuScenes (ablation) | SLIM · KITTI-SF | AEE: 0.1013 -> 0.0668 | Improved |
| train self-supervised on KITTI-RL instead of CARLA (ablation) | SLIM · KITTI-SF | AEE: 0.1454 -> 0.0668 | Improved |
| train baseline on KITTI-RL instead of FT3D (ablation) | PointPWCNet · KITTI-SF | AEE: 0.3569 -> 0.3712 | Hurt |
| self-supervised SLIM vs supervised PointPWCNet on KITTI-SF | PointPWCNet · KITTI-SF | AEE: 0.2578 -> 0.0668 | Improved |
| SLIM vs PointPWCNet on nuScenes self-supervised | PointPWCNet · nuScenes (self-supervised) | AEE 50-50: 0.2756 -> 0.0987 | Improved |
Later papers that compared with this method
3 later accepted papers report SLIM as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.
- FlexPara: Flexible Neural Surface Parameterization (TPAMI 2025 · better on 4 of 4 shared results)
- TARS: Traffic-Aware Radar Scene Flow Estimation (ICCV 2025 · better on 2 of 2 shared results)
- Flatten Anything: Unsupervised Neural Surface Parameterization (NeurIPS 2024 · better on 8 of 8 shared results)
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.
- PointPWCNet — PointPWC-Net: Cost Volume on Point Clouds for (Self-)Supervised Scene Flow Estimation (ECCV 2020) · better on 9 of 9 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
- SeFlow: A Self-Supervised Scene Flow Method in Autonomous Driving (ECCV 2024 · similarity 0.91)
- Motion Cues from Image-based Point Tracking for LiDAR Scene Flow Estimation (NeurIPS 2026 · similarity 0.90)
- 3DSFLabelling: Boosting 3D Scene Flow Estimation by Pseudo Auto-labelling (CVPR 2024 · similarity 0.90)
- ICP-Flow: LiDAR Scene Flow Estimation with ICP (CVPR 2024 · similarity 0.90)
- VoteFlow: Enforcing Local Rigidity in Self-Supervised Scene Flow (CVPR 2025 · similarity 0.90)
- Self-Supervised Class-Agnostic Motion Prediction with Spatial and Temporal Consistency Regularizations (CVPR 2024 · similarity 0.88)
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.
