Research map / Point clouds / Point cloud classification
SCOOP: Self-Supervised Correspondence and Optimization-Based Scene Flow
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 run-time flow refinement (ablation) | SCOOP+ · KITTI_o refinement experiment | EPE: 0.139 -> 0.047 | Helped |
| Add run-time flow refinement (ablation) | FLOT · KITTI_o refinement experiment | EPE: 0.142 -> 0.048 | Helped |
| Restrict correspondence candidates to target points with highest transport mass (ablation) | SCOOP · KITTI_t | EPE: 0.047 -> 0.039 | Helped |
| Add matching confidence weighting (ablation) | SCOOP · KITTI_t | EPE: 0.044 -> 0.039 | Helped |
| Add flow smoothness loss (ablation) | SCOOP · KITTI_t | EPE: 0.056 -> 0.039 | Helped |
| Add run-time flow refinement (ablation) | SCOOP · KITTI_t | EPE: 0.115 -> 0.039 | Helped |
| Train on 1,800 rather than 180 FT3D_o examples (ablation) | SCOOP · KITTI_o | EPE: 0.057 -> 0.047 | Helped |
| Increase training data from 1,800 to 18,000 FT3D_o examples (ablation) | SCOOP · KITTI_o | EPE: 0.047 -> 0.047 | No clear effect |
| Turn off Sinkhorn normalization (ablation) | SCOOP · KITTI_t | EPE: 0.039 -> 0.042 | No clear effect |
| Increase Sinkhorn iterations from 1 to 3 (ablation) | SCOOP · KITTI_t | EPE: 0.039 -> 0.040 | No clear effect |
| Use linear confidence normalization (ablation) | SCOOP · KITTI_t | EPE: 0.039 -> 0.040 | No clear effect |
| Use SCOOP rather than BiPFN, with fewer training examples and self-supervision | KITTI_o · KITTI_o | EPE: 0.065 -> 0.063 | Helped |
Closest papers by meaning
- Self-Point-Flow: Self-Supervised Scene Flow Estimation From Point Clouds With Optimal Transport and Random Walk (CVPR 2021 · similarity 0.93)
- ZeroFlow: Scalable Scene Flow via Distillation (ICLR 2024 · similarity 0.88)
- ICP-Flow: LiDAR Scene Flow Estimation with ICP (CVPR 2024 · similarity 0.87)
- Motion Cues from Image-based Point Tracking for LiDAR Scene Flow Estimation (NeurIPS 2026 · similarity 0.87)
- Rectified Point Flow: Generic Point Cloud Pose Estimation (NeurIPS 2025 · similarity 0.87)
- Zero-Shot Monocular Scene Flow Estimation in the Wild (CVPR 2025 · 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.