Research map / Point clouds / Segmentation and detection
Scribble-Supervised LiDAR Semantic Segmentation
Check 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 the full scribble-supervised pipeline | Cylinder3D · SemanticKITTI valid | mIoU: 57.0 -> 61.3 | Helped |
| Add the full scribble-supervised pipeline | MinkowskiNet · SemanticKITTI valid | mIoU: 55.0 -> 58.5 | Helped |
| Add the full scribble-supervised pipeline | SPVCNN · SemanticKITTI valid | mIoU: 56.9 -> 60.8 | Helped |
| Scribble-label all frames instead of fully labeling 10% of frames | Cylinder3D · SemanticKITTI valid | mIoU: 46.8 -> 57.0 | Helped |
| Add mean teacher consistency loss (ablation) | Scribble-supervised Cylinder3D · SemanticKITTI valid | mIoU: 57.0 -> 59.3 | Helped |
| Add mean teacher consistency loss (ablation) | Scribble-supervised Cylinder3D · SemanticKITTI train | mIoU: 77.6 -> 78.0 | No clear effect |
| Add CRB self-training to mean teacher (ablation) | Scribble-supervised Cylinder3D · SemanticKITTI valid | mIoU: 59.3 -> 60.6 | Helped |
| Add PLS augmentation to mean teacher (ablation) | Scribble-supervised Cylinder3D · SemanticKITTI train | mIoU: +8.0% | Helped |
| Add PLS augmentation to mean teacher (ablation) | Scribble-supervised Cylinder3D · Pseudo-label accuracy | accuracy: 98.1% to 99.0% | Helped |
| Balance pseudo-label selection by range as well as class (ablation) | Class-balanced pseudo-labeling · SemanticKITTI valid | mIoU: +0.5% | Helped |
| Replace CRB with naive pseudo-label sampling (ablation) | Scribble-supervised Cylinder3D · SemanticKITTI valid | mIoU: 59.4 vs 61.3 | Hurt |
| Replace CRB with threshold-based sampling (ablation) | Scribble-supervised Cylinder3D · SemanticKITTI valid | mIoU: 59.1 vs 61.3 | Hurt |
Closest papers by meaning
- LaserMix for Semi-Supervised LiDAR Semantic Segmentation (CVPR 2023 · similarity 0.92)
- Annotator: A Generic Active Learning Baseline for LiDAR Semantic Segmentation (NeurIPS 2023 · similarity 0.90)
- Exploring Scene Affinity for Semi-Supervised LiDAR Semantic Segmentation (CVPR 2025 · similarity 0.88)
- Beyond the Label Itself: Latent Labels Enhance Semi-supervised Point Cloud Panoptic Segmentation (AAAI 2024 · similarity 0.87)
- HiLoTs: High-Low Temporal Sensitive Representation Learning for Semi-Supervised LiDAR Segmentation in Autonomous Driving (CVPR 2025 · similarity 0.86)
- LiSA: LiDAR Localization with Semantic Awareness (CVPR 2024 · 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.