5-way 5-shot Classification on mini-ImageNet (test)
70.44Accuracyamortized Bayesian prototype meta-learning
Evaluation Results
| Method | Links | |
|---|---|---|
| amortized Bayesian prototype meta-learning2022.03 | 70.44 | |
| HyperShotfinetuning=true2022.03 | 69.62 | |
| HyperShotfinetuning=false2022.03 | 68.78 | |
| VSM2022.03 | 68.01 | |
| VERSA2022.03 | 67.37 | |
| OVE PG GP + Cosine (PL)Similarity=Cosine, Protocol=PL2022.03 | 67.14 | |
| SimpleShot2022.03 | 66.92 | |
| Baseline++2022.03 | 66.18 | |
| Reptile2022.03 | 65.99 | |
| R2-D22022.03 | 65.5 | |
| Meta-Mixture2022.03 | 64.6 | |
| OVE PG GP + Cosine (ML)Similarity=Cosine, Protocol=ML2022.03 | 64.58 | |
| VAMPIRE2022.03 | 64.31 | |
| Bayesian MAML2022.03 | 64.23 | |
| RelationNet2022.03 | 64.2 | |
| ProtoNet2022.03 | 64.07 | |
| DKT + BNCosSimSimilarity=BNCosSim2022.03 | 64 | |
| GPLDLA2022.03 | 62.96 | |
| DKT + CosSimSimilarity=Cosine2022.03 | 62.85 | |
| MatchingNet2022.03 | 62.71 | |
| MAML2022.03 | 61.58 | |
| ML-LSTM2022.03 | 60.6 | |
| Feature Transfer2022.03 | 60.51 | |
| Amortized VI2022.03 | 55.68 | |
| SNAIL2022.03 | 55.2 |