5-way Few-shot Classification on SUN (test)
821-shot AccuracyΔ-encoder
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Δ-encoderFeatures=ImageNet model features (trained on disjoint categories)2018.06 | 82 | 93 | |
| Prototypical NetworksFeatures=ImageNet model features (trained on disjoint categories)2018.06 | 74.7 | 94.8 | |
| Nearest neighbor (baseline)Features=ImageNet model features (trained on disjoint categories)2018.06 | 72.7 | 86.7 |