5-way Image Classification on CUB (test)
0.93ECEProtoNet
Evaluation results
| Method | Links | |
|---|---|---|
| ProtoNetshots=5-shot2019.10 | 0.93 | |
| ProtoNetshots=1-shot2019.10 | 1.07 | |
| MAMLshots=1-shot2019.10 | 1.14 | |
| DKT + BNCosSimshots=5-shot2019.10 | 1.15 | |
| Baseline++shots=5-shot2019.10 | 2.04 | |
| MatchingNetshots=5-shot2019.10 | 2.23 | |
| MAMLshots=5-shot2019.10 | 2.47 | |
| DKT + BNCosSimshots=1-shot2019.10 | 2.62 | |
| RelationNetshots=5-shot2019.10 | 2.8 | |
| MatchingNetshots=1-shot2019.10 | 3.11 | |
| RelationNetshots=1-shot2019.10 | 4.13 | |
| Baseline++shots=1-shot2019.10 | 4.91 | |
| Feature Transfershots=1-shot2019.10 | 12.57 | |
| Feature Transfershots=5-shot2019.10 | 18.43 |