Semi-supervised 5-way Classification Accuracy on miniImageNet (test)
84.4AccuracyMCT (Instance)
Evaluation results
| Method | Links | |
|---|---|---|
| MCT (Instance)Backbone=ResNet-12, Shots=5-shot2020.02 | 84.4 | |
| MCT (Instance)Backbone=ResNet-12, Shots=5-shot, Distractors=3 distracting classes2020.02 | 81.3 | |
| LSTBackbone=ResNet-12, Shots=5-shot2020.02 | 78.7 | |
| LSTBackbone=ResNet-12, Shots=5-shot, Distractors=3 distracting classes2020.02 | 77.4 | |
| TPNBackbone=ResNet-12, Shots=5-shot2020.02 | 74.2 | |
| MCT (Instance)Backbone=ResNet-12, Shots=1-shot2020.02 | 73.8 | |
| Masked Soft k-MeansBackbone=ResNet-12, Shots=5-shot2020.02 | 73.6 | |
| TPNBackbone=ResNet-12, Shots=5-shot, Distractors=3 distracting classes2020.02 | 72.4 | |
| Masked Soft k-MeansBackbone=ResNet-12, Shots=5-shot, Distractors=3 distracting classes2020.02 | 72 | |
| LSTBackbone=ResNet-12, Shots=1-shot2020.02 | 70.1 | |
| MCT (Instance)Backbone=ResNet-12, Shots=1-shot, Distractors=3 distracting classes2020.02 | 69.6 | |
| LSTBackbone=ResNet-12, Shots=1-shot, Distractors=3 distracting classes2020.02 | 64.1 | |
| TPNBackbone=ResNet-12, Shots=1-shot2020.02 | 62.7 | |
| Masked Soft k-MeansBackbone=ResNet-12, Shots=1-shot2020.02 | 62.1 | |
| TPNBackbone=ResNet-12, Shots=1-shot, Distractors=3 distracting classes2020.02 | 61.3 | |
| Masked Soft k-MeansBackbone=ResNet-12, Shots=1-shot, Distractors=3 distracting classes2020.02 | 61 |