Few-shot classification on CropDiseases
88.24Accuracy (5-way 5-shot)PsCo
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| PsCoLearning Paradigm=Unsupervised meta-learning, Backbone=Conv5, Meta-training dataset=miniImageNet, Number of evaluation tasks=20002023.03 | 88.24 | — | 94.95 | |
| MoCo v2Learning Paradigm=Self-supervised learning, Backbone=Conv5, Meta-training dataset=miniImageNet, Number of evaluation tasks=20002023.03 | 80.96 | — | 89.85 | |
| SwAVLearning Paradigm=Self-supervised learning, Backbone=Conv5, Meta-training dataset=miniImageNet, Number of evaluation tasks=20002023.03 | 80.15 | — | 89.24 | |
| SimCLRLearning Paradigm=Self-supervised learning, Backbone=Conv5, Meta-training dataset=miniImageNet, Number of evaluation tasks=20002023.03 | 79.9 | — | 88.73 | |
| MAMLLearning Paradigm=Supervised meta-learning, Backbone=Conv5, Meta-training dataset=miniImageNet, Number of evaluation tasks=20002023.03 | 77.76 | — | 83.24 | |
| ProtoNetsLearning Paradigm=Supervised meta-learning, Backbone=Conv5, Meta-training dataset=miniImageNet, Number of evaluation tasks=20002023.03 | 76.01 | — | 83.64 | |
| Meta-GMVAELearning Paradigm=Unsupervised meta-learning, Backbone=Conv5, Meta-training dataset=miniImageNet, Number of evaluation tasks=20002023.03 | 73.56 | — | 81.22 | |
| Meta-SVEBMLearning Paradigm=Unsupervised meta-learning, Backbone=Conv5, Meta-training dataset=miniImageNet, Number of evaluation tasks=2002023.03 | 71.82 | — | 83.13 | |
| BYOLBackbone=ResNet-50, Evaluation protocol=5-way 5-shot2023.03 | — | 80.82 | — | |
| MoCo v2Backbone=ResNet-18, Evaluation protocol=5-way 5-shot2023.03 | — | 81.48 | — | |
| MoCo v2Backbone=ResNet-50, Evaluation protocol=5-way 5-shot2023.03 | — | 82.89 | — | |
| PsCoBackbone=ResNet-18, Base Method=MoCo v2, Evaluation protocol=5-way 5-shot2023.03 | — | 87.91 | — | |
| PsCoBackbone=ResNet-50, Base Method=MoCo v2, Evaluation protocol=5-way 5-shot2023.03 | — | 89.85 | — | |
| PsCoBackbone=ResNet-50, Base Method=BYOL, Evaluation protocol=5-way 5-shot2023.03 | — | 92.92 | — | |
| SupervisedBackbone=ResNet-50, Evaluation protocol=5-way 5-shot2023.03 | — | 90.96 | — |