Few-Shot Image Classification on tiered-ImageNet
79.671-Shot AccOurs-Cosine
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Ours-Cosinebackbone=ViT-S2022.03 | 79.67 | 89.27 | |
| Ours-Classifierbackbone=ViT-S2022.03 | 79.57 | 91.72 | |
| DCbackbone=WRN-28-102022.03 | 78.19 | 89.9 | |
| MN + MCbackbone=ResNet-122022.03 | 74.58 | 86.73 | |
| METAQDAbackbone=WRN2022.03 | 74.29 | 89.41 | |
| CSEIbackbone=ResNet-122022.03 | 73.76 | 87.83 | |
| S2M2_RBackbone=WRN-28-10, N-way=5-way2019.07 | 73.71 | 88.59 | |
| COSOCbackbone=ResNet-122022.03 | 73.57 | 87.57 | |
| CNLbackbone=ResNet-122022.03 | 73.42 | 87.72 | |
| RotationBackbone=WRN-28-10, N-way=5-way2019.07 | 73.04 | 87.89 | |
| PALbackbone=ResNet-122022.03 | 72.25 | 86.95 | |
| TPMNbackbone=ResNet-122022.03 | 72.24 | 86.55 | |
| IEbackbone=ResNet-122022.03 | 72.21 | 87.08 | |
| MELRbackbone=ResNet-122022.03 | 72.14 | 87.01 | |
| DMFbackbone=ResNet-122022.03 | 71.89 | 85.96 | |
| DeepEMDbackbone=ResNet-122022.03 | 71.16 | 86.03 | |
| BMLbackbone=ResNet-122022.03 | 68.99 | 85.49 | |
| Manifold MixupBackbone=WRN-28-10, N-way=5-way2019.07 | 68.19 | 84.61 | |
| LEOBackbone=WRN-28-10, N-way=5-way2019.07 | 66.33 | 81.44 | |
| Baseline-Cosinebackbone=ViT-S2022.03 | 66.04 | 78.05 | |
| DCOBackbone=WRN-28-10, N-way=5-way2019.07 | 65.99 | 81.56 | |
| Baseline++Backbone=WRN-28-10, N-way=5-way2019.07 | 60.98 | 75.93 | |
| RelationNetBackbone=WRN-28-10, N-way=5-way2019.07 | 54.48 | 71.32 | |
| ProtoNetBackbone=WRN-28-10, N-way=5-way2019.07 | 53.31 | 72.69 | |
| MAMLBackbone=WRN-28-10, N-way=5-way2019.07 | 51.67 | 70.3 |