Table 2 Compare the classification accuracy (%) of different methods on source domain test set.
From: Cross-domain few-shot learning based on pseudo-Siamese neural network
Methods | mini-ImageNet | |
|---|---|---|
5-way 1-shot | 5-way 5-shot | |
Delta-encoder39 | 58.7 | 73.6 |
MatchingNet19 | 43.44 ± 0.77 | 55.31 ± 0.73 |
ProtoNet6 | 49.42 ± 0.78 | 68.20 ± 0.66 |
RelationNet4 | 50.44 ± 0.82 | 65.32 ± 0.70 |
TADAM40 | 58.5 ± 0.3 | 76.7 ± 0.3 |
MAML3 | 48.70 ± 1.75 | 63.11 ± 0.92 |
MetaOpt5 | 62.64 ± 0.35 | 78.63 ± 0.68 |
CTM41 | 64.12 ± 0.82 | 80.51 ± 0.13 |
LGM-Net42 | 69.13 + 0.35 | 71.18 + 0.68 |
ResNet-1838 | 59.88 ± 0.88 | 75.71 ± 0.65 |
MTL (ResNet-18)25 | 61.7 ± 1.8 | 75.6 ± 0.9 |
IGN10 | - | 79.94 ± 1.05 |
CDPSN (with branch FC layers) | 58.65 ± 0.92 | 80.68 ± 0.66 |
CDPSN (without branch FC layers) | 60.84 ± 0.97 | 77.69 ± 0.68 |
