Point Cloud Classification on ScanObjectNN Object & Background 1.0
93.29AccuracyACT
Evaluation results
| Method | Links | |
|---|---|---|
| ACTTransfer Protocol=FULL, #Params(M)=22.1, DA=true, Training Augmentation=Rotation2022.12 | 93.29 | |
| ACTTransfer Protocol=FULL, #Params(M)=22.1, DA=false, Training Augmentation=None2022.12 | 91.22 | |
| Point-MAETransfer Protocol=FULL, #Params(M)=22.1, DA=true2022.12 | 90.02 | |
| MaskPointTransfer Protocol=FULL, #Params(M)=22.1, DA=true2022.12 | 89.3 | |
| Point-BERTTransfer Protocol=FULL, #Params(M)=22.1, DA=true2022.12 | 87.43 | |
| PointCNNTransfer Protocol=Supervised Learning Only, #Params(M)=0.6, DA=true2022.12 | 86.1 | |
| OcCoTransfer Protocol=FULL, #Params(M)=22.1, DA=true2022.12 | 84.85 | |
| DGCNNTransfer Protocol=Supervised Learning Only, #Params(M)=1.8, DA=true2022.12 | 82.8 | |
| PointNet++Transfer Protocol=Supervised Learning Only, #Params(M)=1.5, DA=true2022.12 | 82.3 | |
| TransformerTransfer Protocol=FULL, #Params(M)=22.1, DA=true2022.12 | 79.86 | |
| SpiderCNNTransfer Protocol=Supervised Learning Only, DA=true2022.12 | 77.1 | |
| PointNetTransfer Protocol=Supervised Learning Only, #Params(M)=3.5, DA=true2022.12 | 73.3 | |
| MetaSetsBackbone=DGCNN2022.04 | 59.08 | |
| DGCNN2022.04 | 54.42 | |
| LDGCNN2022.04 | 52.5 | |
| ConvPoint2022.04 | 50.67 | |
| MetaSetsBackbone=PointNet2022.04 | 49.5 | |
| PointNet2022.04 | 45.5 | |
| PointCNN2022.04 | 43.92 | |
| PointNet++2022.04 | 43.25 | |
| PointDANBackbone=PointNet2022.04 | 43 |