Few-shot Classification on CUB200 5-way 1-shot
95.8AccuracyPTMAP-SF-SOTp
Evaluation Results
| Method | Links | |
|---|---|---|
| PTMAP-SF-SOTpSOT insertion=without network retraining2022.04 | 95.8 | |
| PTMAP-SF(S)2022.04 | 95.45 | |
| Sill-Net-SOTpSOT insertion=without network retraining2022.04 | 94.86 | |
| Sill-Net(S)2022.04 | 94.73 | |
| Our Method (Improved EM)Setting=Transductive2021.08 | 93.78 | |
| Conference VersionSetting=Transductive2021.08 | 93.2 | |
| Our Method (EM)Setting=Transductive2021.08 | 92.35 | |
| PTMAP-SOTpSOT insertion=without network retraining2022.04 | 91.9 | |
| PTMAP(S)2022.04 | 91.55 | |
| Our MethodSetting=Inductive2021.08 | 88.99 | |
| ICISetting=Transductive2021.08 | 87.87 | |
| BD-CSPNSetting=Transductive2021.08 | 84.9 | |
| RAP-LaplacianShotSetting=Transductive2021.08 | 83.59 | |
| EPNetSetting=Transductive2021.08 | 82.85 | |
| TIM-GDSetting=Transductive2021.08 | 82.2 | |
| LaplacianShotSetting=Transductive2021.08 | 80.96 | |
| CPDESetting=Inductive2021.08 | 80.11 | |
| DCSetting=Inductive2021.08 | 77.22 | |
| SRestoreNetSetting=Transductive2021.08 | 76.85 | |
| MultiSemSetting=Inductive2021.08 | 76.1 | |
| RAP-ProtoNetSetting=Inductive2021.08 | 75.17 | |
| RestoreNetSetting=Inductive2021.08 | 74.32 | |
| CentAlignSetting=Inductive2021.08 | 74.22 | |
| CFASetting=Inductive2021.08 | 73.9 | |
| Neg-CosineSetting=Inductive2021.08 | 72.66 | |
| MAMLSetting=Inductive2021.08 | 55.92 | |
| SimSiam + AugSelfBackbone=ResNet-50, Pre-training Dataset=ImageNet1002021.11 | 48.08 | |
| Supervised + AugSelfBackbone=ResNet-50, Pre-training Dataset=ImageNet1002021.11 | 47.58 | |
| SupervisedBackbone=ResNet-50, Pre-training Dataset=ImageNet1002021.11 | 46.57 | |
| SimSiamBackbone=ResNet-50, Pre-training Dataset=ImageNet1002021.11 | 45.56 | |
| MoCo v2 + AugSelfBackbone=ResNet-50, Pre-training Dataset=ImageNet1002021.11 | 44.17 | |
| MoCo v2Backbone=ResNet-50, Pre-training Dataset=ImageNet1002021.11 | 41.67 | |
| SimSiam + AugSelfBackbone=ResNet-18, Pre-training Dataset=STL102021.11 | 41.6 | |
| MoCo v2 + AugSelfBackbone=ResNet-18, Pre-training Dataset=STL102021.11 | 38.33 | |
| SimSiamBackbone=ResNet-18, Pre-training Dataset=STL102021.11 | 37.97 | |
| MoCo v2Backbone=ResNet-18, Pre-training Dataset=STL102021.11 | 37.62 |