5-way Few-shot Classification on AWA2 (test)
90.51-shot AccΔ-encoder
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Δ-encoderFeatures=ImageNet model features (trained on disjoint categories)2018.06 | 90.5 | 96.4 | |
| Prototypical NetworksFeatures=ImageNet model features (trained on disjoint categories)2018.06 | 80.8 | 95.3 | |
| Nearest neighbor (baseline)Features=ImageNet model features (trained on disjoint categories)2018.06 | 65.9 | 84.2 |