Few-Shot Image Classification on tiered-ImageNet (test)
54.15-shot AccuracyUNEM-Gaussian
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| UNEM-GaussianBackbone=WRN28-102024.12 | 54.1 | 66.8 | 74.7 | |
| UNEM-GaussianBackbone=ResNet-182024.12 | 52.3 | 65.7 | 73.2 | |
| α-TIMBackbone=WRN28-102024.12 | 45.8 | 51.4 | 55.2 | |
| PADDLEBackbone=ResNet-182024.12 | 45.4 | 61.4 | 70.6 | |
| PADDLEBackbone=WRN28-102024.12 | 43.9 | 59.4 | 69.9 | |
| α-TIMBackbone=ResNet-182024.12 | 43.8 | 48.3 | 51.9 | |
| TIMBackbone=WRN28-102024.12 | 36.1 | 39 | 38.5 | |
| BaselineBackbone=WRN28-102024.12 | 31.9 | 39 | 45.6 | |
| LaplacianShotBackbone=WRN28-102024.12 | 31.4 | 37.3 | 41.5 | |
| BaselineBackbone=ResNet-182024.12 | 29.7 | 36.3 | 42.2 | |
| LaplacianShotBackbone=ResNet-182024.12 | 29.6 | 35.4 | 39.1 | |
| TIMBackbone=ResNet-182024.12 | 29.3 | 28.7 | 27.8 | |
| BD-CSPNBackbone=WRN28-102024.12 | 18 | 18.4 | 18.2 | |
| BD-CSPNBackbone=ResNet-182024.12 | 11.4 | 11 | 11.7 | |
| PT-MAPBackbone=ResNet-182024.12 | 5.2 | 6 | 6.6 | |
| PT-MAPBackbone=WRN28-102024.12 | 5.2 | 6 | 6.6 |