5-way 20-shot Classification on miniImageNet (test)
0.71035-way 20-shot AccuracyMAML
Evaluation Results
| Method | Links | |
|---|---|---|
| MAMLClustering=N/A, Training Protocol=Supervised (Upper Bound), Backbone=4-layer CNN2020.04 | 0.7103 | |
| ProtoNetsClustering=N/A, Training Protocol=Supervised (Upper Bound), Backbone=4-layer CNN2020.04 | 0.7005 | |
| ULDA-ProtoNetsClustering=N/A, Training Protocol=Unsupervised, Backbone=4-layer CNN2020.04 | 0.6431 | |
| CACTUs-MAMLClustering=DeepCluster, Training Protocol=Unsupervised, Backbone=4-layer CNN2020.04 | 0.6384 | |
| ULDA-MetaOptNetClustering=N/A, Training Protocol=Unsupervised, Backbone=4-layer CNN2020.04 | 0.6358 | |
| CACTUs-ProtoNetsClustering=DeepCluster, Training Protocol=Unsupervised, Backbone=4-layer CNN2020.04 | 0.6154 | |
| UMTRAClustering=N/A, Training Protocol=Unsupervised, Backbone=4-layer CNN2020.04 | 0.6111 | |
| knn-nearest neighborsClustering=DeepCluster, Training Protocol=Unsupervised, Backbone=4-layer CNN2020.04 | 0.5644 | |
| linear classifierClustering=DeepCluster, Training Protocol=Unsupervised, Backbone=4-layer CNN2020.04 | 0.5619 | |
| MLP with dropoutClustering=DeepCluster, Training Protocol=Unsupervised, Backbone=4-layer CNN2020.04 | 0.5271 | |
| Training from scratchClustering=N/A, Training Protocol=Unsupervised, Backbone=4-layer CNN2020.04 | 0.5153 | |
| cluster matchingClustering=DeepCluster, Training Protocol=Unsupervised, Backbone=4-layer CNN2020.04 | 0.2497 |