Research map / Point clouds / Point cloud classification
CrossPoint: Self-Supervised Cross-Modal Contrastive Learning for 3D Point Cloud Understanding
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 |
|---|---|---|---|
| combine intra-modal point-cloud contrastive learning with cross-modal image-point-cloud alignment (ablation) | CrossPoint · ModelNet40 | linear classification accuracy: +1.2% | Helped |
| combine intra-modal point-cloud contrastive learning with cross-modal image-point-cloud alignment (ablation) | CrossPoint · ScanObjectNN | linear classification accuracy: +0.7% | Helped |
| use 2 rendered 2D images instead of 1 (ablation) | CrossPoint · ModelNet40 | linear accuracy: 91.2 -> 91.2 | No clear effect |
| use 3 rendered 2D images instead of 1 (ablation) | CrossPoint · ModelNet40 | linear accuracy: 91.2 -> 90.9 | Hurt |
| use 4 rendered 2D images instead of 1 (ablation) | CrossPoint · ModelNet40 | linear accuracy: 91.2 -> 91.0 | Hurt |
| use 5 rendered 2D images instead of 1 (ablation) | CrossPoint · ModelNet40 | linear accuracy: 91.2 -> 90.5 | Hurt |
| fine-tune CrossPoint image backbone with RFS | ResNet-50 image backbone · CIFAR-FS 5-way 1-shot | accuracy: 60.20±0.87 -> 64.45±0.86 | Helped |
| fine-tune CrossPoint image backbone with RFS | ResNet-50 image backbone · CIFAR-FS 5-way 5-shot | accuracy: 76.79±0.71 -> 80.14±0.65 | Helped |
| use the pretrained CrossPoint image backbone without supervised fine-tuning | ResNet-50 image backbone · CIFAR-FS | few-shot image classification accuracy | Hurt |
Closest papers by meaning
- MM-Point: Multi-View Information-Enhanced Multi-Modal Self-Supervised 3D Point Cloud Understanding (AAAI 2024 · similarity 0.90)
- Sonata: Self-Supervised Learning of Reliable Point Representations (CVPR 2025 · similarity 0.89)
- FreeReg: Image-to-Point Cloud Registration Leveraging Pretrained Diffusion Models and Monocular Depth Estimators (ICLR 2024 · similarity 0.88)
- FreePoint: Unsupervised Point Cloud Instance Segmentation (CVPR 2024 · similarity 0.88)
- Interpretable3D: An Ad-Hoc Interpretable Classifier for 3D Point Clouds (AAAI 2024 · similarity 0.87)
- Mitigating Object Dependencies: Improving Point Cloud Self-Supervised Learning through Object Exchange (CVPR 2024 · 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.