Research map / Point clouds / Point cloud classification
Interpretable point cloud classification using multiple instance learning
Matt De Vries, Reed Naidoo, Olga Fourkioti, Lucas G. Dent, Nathan Curry, Chris Dunsby, Chris Bakal
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 MIL pooling to backbone | DGCNN · IntrA | mACC: 90.6 -> 95.2 | Improved |
| add MIL pooling to backbone | PointNet · IntrA | mACC: 81.8 -> 82.0 | No clear effect |
| add MIL pooling to backbone | Transformer · IntrA | mACC: +5.1 | Improved |
| add MIL pooling to backbone | PointNet · ScanObjectNN | oACC: 68.2 -> 73.8 | Improved |
| add MIL pooling to backbone | DGCNN · ScanObjectNN | oACC: 86.0 -> 85.6 | No clear effect |
| add MIL pooling to backbone | PointMLPElite · RBC | mACC: 84.7 -> 92.1 | Improved |
| add MIL pooling to backbone | PointNet · ModelNet40 | oACC: 89.2 -> 90.7 | Improved |
| use Additive MIL pooling instead of CLAIM (ablation) | Transformer · IntrA | AOPCR: 14.023 -> 18.162 | Improved |
| use Additive MIL pooling instead of CLAIM (ablation) | DGCNN · IntrA | AOPCR: 6.033 -> 4.486 | Hurt |
| use Additive instead of Attention pooling (ablation) | Transformer · IntrA | AOPCR: 14.541 -> 18.162 | Improved |
| use Conjunctive instead of Attention pooling (ablation) | Transformer · IntrA | NDCG@n: 0.539 -> 0.610 | Improved |
| add contextual attention (k-NN smoothing, k>0) (ablation) | Attention, Additive, Conjunctive pooling · IntrA | F1 and mACC: > 97% | Improved |
Closest papers by meaning
- Interpretable3D: An Ad-Hoc Interpretable Classifier for 3D Point Clouds (AAAI 2024 · similarity 0.86)
- Symb-xMIL: Symbolic Explanations for Multiple Instance Learning in Digital Pathology (NeurIPS 2026 · similarity 0.84)
- MorphoHELM: A Comprehensive Benchmark for Evaluating Representations for Microscopy-Based Morphology Assays (NeurIPS 2026 · similarity 0.83)
- Point2SSM: Learning Morphological Variations of Anatomies from Point Clouds (ICLR 2024 · similarity 0.83)
- xMIL: Insightful Explanations for Multiple Instance Learning in Histopathology (NeurIPS 2024 · similarity 0.83)
- HiPoNet: A Multi-View Simplicial Complex Network for High Dimensional Point-Cloud and Single-Cell data (NeurIPS 2025 · similarity 0.83)
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.
