Few-shot Classification on ModelNet40
99.2Mean AccuracyMPL-MAE
Evaluation Results
| Method | Links | |
|---|---|---|
| MPL-MAELearning paradigm=with Self-Supervised Representation Learning, N-way=5-way, K-shot=20-shot2026.06 | 99.2 | |
| PCP-MAELearning paradigm=with Self-Supervised Representation Learning, N-way=5-way, K-shot=20-shot2026.06 | 99.1 | |
| Point-FEMAELearning paradigm=with Self-Supervised Representation Learning, N-way=5-way, K-shot=20-shot2026.06 | 98.6 | |
| Point-M2AELearning paradigm=with Self-Supervised Representation Learning, N-way=5-way, K-shot=20-shot2026.06 | 98.3 | |
| Point-MAELearning paradigm=with Self-Supervised Representation Learning, N-way=5-way, K-shot=20-shot2026.06 | 97.8 | |
| MPL-MAELearning paradigm=with Self-Supervised Representation Learning, N-way=5-way, K-shot=10-shot2026.06 | 97.7 | |
| PCP-MAELearning paradigm=with Self-Supervised Representation Learning, N-way=5-way, K-shot=10-shot2026.06 | 97.4 | |
| MaskPointLearning paradigm=with Self-Supervised Representation Learning, N-way=5-way, K-shot=20-shot2026.06 | 97.2 | |
| Point-FEMAELearning paradigm=with Self-Supervised Representation Learning, N-way=5-way, K-shot=10-shot2026.06 | 97.2 | |
| Point-M2AELearning paradigm=with Self-Supervised Representation Learning, N-way=5-way, K-shot=10-shot2026.06 | 96.8 | |
| Point-BERTLearning paradigm=with Self-Supervised Representation Learning, N-way=5-way, K-shot=20-shot2026.06 | 96.3 | |
| Point-MAELearning paradigm=with Self-Supervised Representation Learning, N-way=5-way, K-shot=10-shot2026.06 | 96.3 | |
| MPL-MAELearning paradigm=with Self-Supervised Representation Learning, N-way=10-way, K-shot=20-shot2026.06 | 96.3 | |
| PCP-MAELearning paradigm=with Self-Supervised Representation Learning, N-way=10-way, K-shot=20-shot2026.06 | 95.9 | |
| Point-FEMAELearning paradigm=with Self-Supervised Representation Learning, N-way=10-way, K-shot=20-shot2026.06 | 95.8 | |
| MaskPointLearning paradigm=with Self-Supervised Representation Learning, N-way=5-way, K-shot=10-shot2026.06 | 95 | |
| Point-MAELearning paradigm=with Self-Supervised Representation Learning, N-way=10-way, K-shot=20-shot2026.06 | 95 | |
| Point-M2AELearning paradigm=with Self-Supervised Representation Learning, N-way=10-way, K-shot=20-shot2026.06 | 95 | |
| Point-BERTLearning paradigm=with Self-Supervised Representation Learning, N-way=5-way, K-shot=10-shot2026.06 | 94.6 | |
| MPL-MAELearning paradigm=with Self-Supervised Representation Learning, N-way=10-way, K-shot=10-shot2026.06 | 94.4 | |
| Point-FEMAELearning paradigm=with Self-Supervised Representation Learning, N-way=10-way, K-shot=10-shot2026.06 | 94 | |
| PCP-MAELearning paradigm=with Self-Supervised Representation Learning, N-way=10-way, K-shot=10-shot2026.06 | 93.5 | |
| MaskPointLearning paradigm=with Self-Supervised Representation Learning, N-way=10-way, K-shot=20-shot2026.06 | 93.4 | |
| Point-BERTLearning paradigm=with Self-Supervised Representation Learning, N-way=10-way, K-shot=20-shot2026.06 | 92.7 | |
| Point-MAELearning paradigm=with Self-Supervised Representation Learning, N-way=10-way, K-shot=10-shot2026.06 | 92.6 | |
| OcCoLearning paradigm=Supervised Learning Only, N-way=5-way, K-shot=20-shot2026.06 | 92.5 | |
| Point-M2AELearning paradigm=with Self-Supervised Representation Learning, N-way=10-way, K-shot=10-shot2026.06 | 92.3 | |
| MaskPointLearning paradigm=with Self-Supervised Representation Learning, N-way=10-way, K-shot=10-shot2026.06 | 91.4 | |
| Point-BERTLearning paradigm=with Self-Supervised Representation Learning, N-way=10-way, K-shot=10-shot2026.06 | 91 | |
| Point-NNway=5-way, shot=20-shot2024.12 | 90.9 | |
| Point-GNway=5-way, shot=20-shot2024.12 | 90.9 | |
| Point-GNway=5-way, shot=10-shot2024.12 | 90.7 | |
| OcCoLearning paradigm=Supervised Learning Only, N-way=5-way, K-shot=10-shot2026.06 | 90.6 | |
| Point-NNway=5-way, shot=10-shot2024.12 | 88.8 | |
| OcCoLearning paradigm=Supervised Learning Only, N-way=10-way, K-shot=20-shot2026.06 | 86.5 | |
| Point-GNway=10-way, shot=20-shot2024.12 | 86.4 | |
| Point-NNway=10-way, shot=20-shot2024.12 | 84.9 | |
| OcCoLearning paradigm=Supervised Learning Only, N-way=10-way, K-shot=10-shot2026.06 | 82.9 | |
| Point-GNway=10-way, shot=10-shot2024.12 | 81.6 | |
| Point-NNway=10-way, shot=10-shot2024.12 | 79.9 | |
| PointCNNway=5-way, shot=20-shot2024.12 | 68.6 | |
| 3D-GANway=5-way, shot=20-shot2024.12 | 65.8 | |
| PointCNNway=5-way, shot=10-shot2024.12 | 65.4 | |
| PointNetway=5-way, shot=20-shot2024.12 | 57.8 | |
| PointNetLearning paradigm=Supervised Learning Only, N-way=5-way, K-shot=20-shot2026.06 | 57.8 | |
| 3D-GANway=5-way, shot=10-shot2024.12 | 55.8 | |
| PointNetway=5-way, shot=10-shot2024.12 | 52 | |
| PointNetLearning paradigm=Supervised Learning Only, N-way=5-way, K-shot=10-shot2026.06 | 52 | |
| PointCNNway=10-way, shot=20-shot2024.12 | 50 | |
| 3D-GANway=10-way, shot=20-shot2024.12 | 48.4 | |
| PointNetway=10-way, shot=10-shot2024.12 | 46.6 | |
| PointCNNway=10-way, shot=10-shot2024.12 | 46.6 | |
| PointNetLearning paradigm=Supervised Learning Only, N-way=10-way, K-shot=10-shot2026.06 | 46.6 | |
| PointNet++way=5-way, shot=20-shot2024.12 | 42.4 | |
| DGCNNway=5-way, shot=20-shot2024.12 | 40.8 | |
| DGCNNLearning paradigm=Supervised Learning Only, N-way=5-way, K-shot=20-shot2026.06 | 40.8 | |
| 3D-GANway=10-way, shot=10-shot2024.12 | 40.3 | |
| PointNet++way=5-way, shot=10-shot2024.12 | 38.5 | |
| FoldingNetway=5-way, shot=20-shot2024.12 | 35.8 | |
| PointNetway=10-way, shot=20-shot2024.12 | 35.2 | |
| PointNetLearning paradigm=Supervised Learning Only, N-way=10-way, K-shot=20-shot2026.06 | 35.2 | |
| FoldingNetway=5-way, shot=10-shot2024.12 | 33.4 | |
| DGCNNway=5-way, shot=10-shot2024.12 | 31.6 | |
| DGCNNLearning paradigm=Supervised Learning Only, N-way=5-way, K-shot=10-shot2026.06 | 31.6 | |
| PointNet++way=10-way, shot=10-shot2024.12 | 23 | |
| DGCNNway=10-way, shot=10-shot2024.12 | 19.9 | |
| DGCNNLearning paradigm=Supervised Learning Only, N-way=10-way, K-shot=10-shot2026.06 | 19.9 | |
| PointNet++way=10-way, shot=20-shot2024.12 | 18.8 | |
| FoldingNetway=10-way, shot=10-shot2024.12 | 18.6 | |
| DGCNNway=10-way, shot=20-shot2024.12 | 16.9 | |
| DGCNNLearning paradigm=Supervised Learning Only, N-way=10-way, K-shot=20-shot2026.06 | 16.9 | |
| FoldingNetway=10-way, shot=20-shot2024.12 | 15.4 |