Research map / Point clouds / Point cloud classification
Point-In-Context: Understanding Point Cloud via In-Context Learning
Mengyuan Liu, Zhongbin Fang, Xia Li, Joachim M. Buhmann, Deheng Ye, Xiangtai Li, Chen Change Loy
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 Objaverse data to training (ablation) | PIC-Sep · ShapeNet In-Context: part segmentation | mIoU: 74.95 -> 81.96 | Improved |
| add Objaverse data to training (ablation) | PIC-Cat · ShapeNet In-Context: part segmentation | mIoU: 78.95 -> 80.13 | Improved |
| add Objaverse data to training (ablation) | PIC-Sep · ShapeNet In-Context: registration | Chamfer distance x1000: 10.3 -> 8.2 | Improved |
| add Objaverse data to training (ablation) | PIC-Cat · ShapeNet In-Context: reconstruction | Chamfer distance x1000: 4.3 -> 4.3 | No clear effect |
| add Objaverse data to training (ablation) | PIC-Sep · ModelNet40 registration (unseen) | Chamfer distance x1000: 20.3 -> 15.3 | Improved |
| add Objaverse data to training (ablation) | PIC-Cat · ModelNet40 registration (unseen) | Chamfer distance x1000: 26.9 -> 22.1 | Improved |
| add In-Context Labeling and Enhancing (PIC -> PIC++) | PIC-Sep · ShapeNet In-Context: part segmentation | mIoU: 74.95 -> 85.53 | Improved |
| add In-Context Labeling and Enhancing (PIC -> PIC++) | PIC-Cat · ShapeNet In-Context: part segmentation | mIoU: 78.95 -> 85.27 | Improved |
| add In-Context Labeling and Enhancing (PIC -> PIC++) | PIC-Sep · ShapeNet In-Context: registration | Chamfer distance x1000: 10.3 -> 7.1 | Improved |
| add In-Context Labeling and Enhancing (PIC -> PIC++) | PIC-Cat · ShapeNet In-Context: reconstruction | Chamfer distance x1000: 4.3 -> 4.3 | No clear effect |
| add In-Context Labeling and Enhancing (PIC -> PIC++), trained from scratch | PIC-Cat · Multi-Entity ShapeNetPart (test) | instance mIoU: 64.52 -> 81.13 | Improved |
| add In-Context Labeling and Enhancing (PIC -> PIC++), trained from scratch | PIC-Sep · AKB-48 (one-shot, unseen) | instance mIoU: 19.58 -> 68.96 | Improved |
Closest papers by meaning
- Explore In-Context Learning for 3D Point Cloud Understanding (NeurIPS 2023 · similarity 0.93)
- DG-PIC: Domain Generalized Point-In-Context Learning for Point Cloud Understanding (ECCV 2024 · similarity 0.90)
- Exploring Data-Efficient 3D Scene Understanding With Contrastive Scene Contexts (CVPR 2021 · similarity 0.89)
- Group Contextual Encoding for 3D Point Clouds (NeurIPS 2020 · similarity 0.89)
- Test-Time Optimization of 3D Point Cloud LLM via Manifold-Aware In-Context Guidance and Refinement (ICLR 2026 · similarity 0.89)
- PointGLR: Unsupervised Structural Representation Learning of 3D Point Clouds (TPAMI 2022 · similarity 0.89)
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.
