Few Shot Image Classification On Imagenet 0
Metrics
Accuracy
Results
Performance results of various models on this benchmark
Model Name | Accuracy | Paper Title | Repository |
|---|---|---|---|
| CLIP (ViT B/32) | 63.2% | Learning Transferable Visual Models From Natural Language Supervision | |
| CLIP (ResNet50) | 59.6% | Learning Transferable Visual Models From Natural Language Supervision | |
| DebiasPL (ResNet50) | 68.3% | Debiased Learning from Naturally Imbalanced Pseudo-Labels | |
| Synthesised Classifier | 1.5% | Synthesized Classifiers for Zero-Shot Learning | |
| ConSE | 1.4% | Zero-Shot Learning by Convex Combination of Semantic Embeddings |
0 of 5 row(s) selected.