3D Object Classification on ScanObjectNN PB_T50_RS v1.0
90.63AccuracyReCon
Evaluation Results
| Method | Links | |
|---|---|---|
| ReCon#P (M)=43.6, Input=2k Points2024.08 | 90.63 | |
| PCP-MAE#P (M)=22.1, Input=2k Points2024.08 | 90.35 | |
| Point-FEMAE#P (M)=27.4, Input=2k Points2024.08 | 90.22 | |
| I2P-MAE#P (M)=15.3, Input=2k Points2024.08 | 90.11 | |
| P2P-HorNet#P (M)=195.8, Input=40 Images2024.08 | 89.3 | |
| P2P-HorNetLearning Paradigm=Supervised Learning, Backbone=HorNet2025.12 | 89.3 | |
| PointDicoLearning Paradigm=Diffusion-based SSL, Voting Strategy=true2025.12 | 89.1 | |
| ACT#P (M)=22.1, Input=2k Points2024.08 | 88.21 | |
| SFRInput=20 Images2024.08 | 87.8 | |
| PointNeXtLearning Paradigm=Supervised Learning2025.12 | 87.7 | |
| P2P-RN101Learning Paradigm=Supervised Learning, Backbone=ResNet-1012025.12 | 87.4 | |
| PointGPT#P (M)=19.5, Input=2k Points2024.08 | 86.9 | |
| Point-M2AE#P (M)=15.3, Input=2k Points2024.08 | 86.43 | |
| Joint-MAEInput=2k Points2024.08 | 86.07 | |
| TAP#P (M)=22.1, Input=2k Points2024.08 | 85.67 | |
| PointMLP#P (M)=12.6, Input=1k Points2024.08 | 85.4 | |
| PointMLPLearning Paradigm=Supervised Learning2025.12 | 85.4 | |
| Point-MAE#P (M)=22.1, Input=2k Points2024.08 | 85.18 | |
| MaskPointInput=2k Points2024.08 | 84.3 | |
| Point-BERT#P (M)=22.1, Input=1k Points2024.08 | 83.07 | |
| MVTN#P (M)=11.2, Input=20 Images2024.08 | 82.8 | |
| SimpleViewInput=6 Images2024.08 | 80.5 | |
| SimpleViewLearning Paradigm=Supervised Learning2025.12 | 80.5 | |
| DGCNN#P (M)=1.8, Input=1k Points2024.08 | 78.1 | |
| PointNet++#P (M)=1.5, Input=1k Points2024.08 | 77.9 | |
| PointNet#P (M)=3.5, Input=1k Points2024.08 | 68 |