Few Shot Image Classification On Mini 1
Métriques
Accuracy
Résultats
Résultats de performance de divers modèles sur ce benchmark
Tableau comparatif
| Nom du modèle | Accuracy |
|---|---|
| transductive-few-shot-learning-with-meta | 78.55% |
| prototype-rectification-for-few-shot-learning | 70.31% |
| hypershot-few-shot-learning-by-kernel | 53.18% |
| task-augmentation-by-rotating-for-meta | 65.95% |
| self-supervised-learning-for-few-shot-image | 76.82% |
| task-augmentation-by-rotating-for-meta | 65.38% |
| few-shot-learning-with-global-class | 53.21 |
| negative-margin-matters-understanding-margin | 63.85 |
| embedding-propagation-smoother-manifold-for | 77.27% |
| probabilistic-model-agnostic-meta-learning | 50.13% |
| basetransformers-attention-over-base-data | 70.88% |
| Modèle 12 | 70.0% |
| baby-steps-towards-few-shot-learning-with | 67.2% |
| leveraging-the-feature-distribution-in | 82.92% |
| diversity-with-cooperation-ensemble-methods | 63.73% |
| exploiting-unsupervised-inputs-for-accurate | 76.47% |