Research map / Point clouds / Point cloud classification
A Network Architecture for Point Cloud Classification via Automatic Depth Images Generation
Riccardo Roveri, Lukas Rahmann, Cengiz Oztireli, Markus Gross
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 a second learned depth view instead of using one view (ablation) | Ours · ModelNet40, random 3D rotations | overall accuracy: 0.854 -> 0.869 | Improved |
| Add a second learned depth view instead of using one view (ablation) | Ours · ModelNet40, random y-axis rotations | overall accuracy: 0.873 -> 0.884 | Improved |
| Increase learned depth views from two to four (ablation) | Ours · ModelNet40, random 3D rotations | overall accuracy: 0.869 -> 0.872 | Improved |
| Increase learned depth views from two to four (ablation) | Ours · ModelNet40, random y-axis rotations | overall accuracy: 0.884 -> 0.885 | Improved |
| Use more than four learned views (ablation) | Ours · ModelNet40 | classification accuracy | No clear effect |
| Replace learned view prediction with PCA-based view selection (ablation) | Ours, one-view architecture · ModelNet40, random 3D rotations | overall accuracy: 0.854 -> 0.844 | Hurt |
| Replace learned view prediction with PCA-based view selection (ablation) | Ours, two-view architecture · ModelNet40, random 3D rotations | overall accuracy: 0.869 -> 0.850 | Hurt |
| Replace learned view prediction with random view selection (ablation) | Ours, one-view architecture · ModelNet40, random 3D rotations | overall accuracy: 0.854 -> 0.849 | Hurt |
| Replace learned view prediction with random view selection (ablation) | Ours, two-view architecture · ModelNet40, random 3D rotations | overall accuracy: 0.869 -> 0.859 | Hurt |
| Replace learned view prediction with random view selection (ablation) | Ours, four-view architecture · ModelNet40, random 3D rotations | overall accuracy: 0.872 -> 0.868 | Hurt |
| Replace learned view prediction with random view selection (ablation) | Ours, one-view architecture · ModelNet40, random y-axis rotations | overall accuracy: 0.873 -> 0.849 | Hurt |
| Replace learned view prediction with random view selection (ablation) | Ours, two-view architecture · ModelNet40, random y-axis rotations | overall accuracy: 0.884 -> 0.859 | Hurt |
Closest papers by meaning
- Cloud Transformers: A Universal Approach to Point Cloud Processing Tasks (ICCV 2021 · similarity 0.90)
- Generative PointNet: Deep Energy-Based Learning on Unordered Point Sets for 3D Generation, Reconstruction and Classification (CVPR 2021 · similarity 0.88)
- PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation (CVPR 2017 · similarity 0.88)
- Image2Point: 3D Point-Cloud Understanding with 2D Image Pretrained Models (ECCV 2022 · similarity 0.88)
- Fully-Convolutional Point Networks for Large-Scale Point Clouds (ECCV 2018 · similarity 0.88)
- DeepLA-Net: Very Deep Local Aggregation Networks for Point Cloud Analysis (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.
