Research map / Point clouds / Segmentation and detection
Geometry and Uncertainty-Aware 3D Point Cloud Class-Incremental Semantic Segmentation
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 |
|---|---|---|---|
| Use the proposed GFT and UPG framework instead of LwF | DGCNN incremental segmentation · S3DIS S0, Cnovel=5 | mIoU (%): 40.88 -> 45.33 | Helped |
| Use the proposed GFT and UPG framework instead of LwF | DGCNN incremental segmentation · S3DIS S1, Cnovel=5 | mIoU (%): 41.43 -> 44.72 | Helped |
| Use the proposed GFT and UPG framework instead of LwF | DGCNN incremental segmentation · ScanNet S0, Cnovel=5 | mIoU (%): 26.13 -> 28.98 | Helped |
| Use the proposed GFT and UPG framework instead of LwF | DGCNN incremental segmentation · ScanNet S1, Cnovel=5 | mIoU (%): 27.50 -> 28.41 | Helped |
| Add geometry-aware feature-relation transfer (GFT) (ablation) | Fine-Tuning on S3DIS · S3DIS S1, Cnovel=5 | mIoU (%): 26.53 -> 29.98 | Helped |
| Add uncertainty-aware pseudo-label generation (UPG) (ablation) | Fine-Tuning on S3DIS · S3DIS S1, Cnovel=5 | mIoU (%): 26.53 -> 42.84 | Helped |
| Add both GFT and UPG (ablation) | Fine-Tuning on S3DIS · S3DIS S1, Cnovel=5 | mIoU (%): 26.53 -> 44.72 | Helped |
| Add geometry-aware feature-relation transfer (GFT) (ablation) | Fine-Tuning on ScanNet · ScanNet S0, Cnovel=5 | mIoU (%): 10.45 -> 12.26 | Helped |
| Add uncertainty-aware pseudo-label generation (UPG) (ablation) | Fine-Tuning on ScanNet · ScanNet S0, Cnovel=5 | mIoU (%): 10.45 -> 27.28 | Helped |
| Add both GFT and UPG (ablation) | Fine-Tuning on ScanNet · ScanNet S0, Cnovel=5 | mIoU (%): 10.45 -> 28.98 | Helped |
| Add GFT and UPG to LwF-3D (ablation) | DGCNN incremental classification · ModelNet40 → ScanObjectNN | Acco: 73.34 -> 78.19 | Helped |
| Add GFT and UPG to LwF-3D (ablation) | PointNet incremental classification · ModelNet40 → ScanObjectNN | Acco: 84.77 -> 86.84 | Helped |
Closest papers by meaning
- Novel Class Discovery for 3D Point Cloud Semantic Segmentation (CVPR 2023 · similarity 0.92)
- Few-Shot 3D Point Cloud Semantic Segmentation (CVPR 2021 · similarity 0.91)
- PDF: A Probability-Driven Framework for Open World 3D Point Cloud Semantic Segmentation (CVPR 2024 · similarity 0.91)
- Hyperbolic Uncertainty-Aware Few-Shot Incremental Point Cloud Segmentation (CVPR 2025 · similarity 0.91)
- Geometrically-driven Aggregation for Zero-shot 3D Point Cloud Understanding (CVPR 2024 · similarity 0.90)
- LogoSP: Local-global Grouping of Superpoints for Unsupervised Semantic Segmentation of 3D Point Clouds (CVPR 2025 · similarity 0.90)
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.