Table 2 Performance of different methods on benchmarks
From: A survey of few-shot learning in smart agriculture: developments, applications, and challenges
Method | Model | Backbone | Omniglot | Mini-ImageNet | Tiered-ImageNet | |||
|---|---|---|---|---|---|---|---|---|
1-shot | 5-shot | 1-shot | 5-shot | 1-shot | 5-shot | |||
Data Augmentation | AFHN [27] | ResNet-18 | – | – | 62.38 ± 0.72 | 78.16 ± 0.56 | – | – |
∆-encoder [28] | ResNet-18 | – | – | 59.9 | 69.7 | – | – | |
Metric Learning | MatchingNet [31] | ResNet-12 | 97.9 | 98.7 | 63.08 ± 0.80 | 75.99 ± 0.60 | 68.50 ± 0.92 | 80.60 ± 0.71 |
RelationNet [34] | Conv-4 | 99.6 ± 0.2 | 99.8 ± 0.1 | 50.44 ± 0.82 | 65.32 ± 0.70 | 54.48 ± 0.93 | 71.32 ± 0.78 | |
ProtoNet [32] | ResNet-12 | 98.8 | 99.7 | 60.37 ± 0.83 | 78.02 ± 0.57 | – | – | |
DeepEMD [33] | ResNet-12 | – | – | 65.91 ± 0.82 | 82.41 ± 0.56 | 71.16 ± 0.87 | 86.03 ± 0.58 | |
External Memory | MetaNet [38] | ResNet-12 | 99.9 | – | 49.21 ± 0.96 | – | – | – |
MMNet [39] | CNN + LSTM | 99.28 ± 0.08 | 99.77 ± 0.1 | 53.37 ± 0.48 | 66.97 ± 0.35 | – | – | |
[40] | ResNet-10 | – | – | 55.45 ± 0.89% | 70.13 ± 0.68% | – | – | |
Parameter Optimization | MAML [41] | Conv-4 | 98.7 ± 0.4 | 99.9 ± 0.1 | 48.70 ± 1.75 | 63.11 ± 0.92 | – | – |
Reptile [43] | Conv-4 | 95.39 ± 0.09 | 98.90 ± 0.1 | 47.07 ± 0.26% | 62.74 ± 0.37% | – | – | |
MetaNAS [45] | CNN | – | – | 63.1 ± 0.3 | 79.5 ± 0.2 | – | – | |