Point Cloud Classification on ModelNet40 point perturbation attack
86.99AccuracyIF-Defense
Evaluation results
| Method | Links | |
|---|---|---|
| IF-DefenseModel=PointNet++2021.01 | 86.99 | |
| PointCutMix-RModel=PointNet++2021.01 | 86.71 | |
| IF-DefenseModel=DGCNN2021.01 | 85.53 | |
| PointCutMix-KModel=PointNet++2021.01 | 84.93 | |
| PointCutMix-RModel=DGCNN2021.01 | 83.14 | |
| DUP-NetModel=PointNet++2021.01 | 80.63 | |
| SORModel=PointNet++2021.01 | 77.67 | |
| PointCutMix-KModel=DGCNN2021.01 | 76.69 | |
| SORModel=DGCNN2021.01 | 76.5 | |
| SRSModel=PointNet++2021.01 | 73.14 | |
| SRSModel=DGCNN2021.01 | 50.2 | |
| DUP-NetModel=DGCNN2021.01 | 42.67 | |
| No DefenseModel=PointNet++2021.01 | 0 | |
| No DefenseModel=DGCNN2021.01 | 0 |