Image Classification on 11 Classification Datasets Average 8-shot
85.3Accuracy2SFS
Evaluation Results
| Method | Links | |
|---|---|---|
| 2SFSBackbone=ViT-L/142025.03 | 85.3 | |
| SigLIP2Classifier=LDA, Embedding Projection=PCA<-, Backbone=ViT-B2026.03 | 84.7 | |
| SigLIP2Classifier=LDA, Embedding Projection=Original, Backbone=ViT-B2026.03 | 84.5 | |
| SigLIPClassifier=LDA, Embedding Projection=Original, Backbone=ViT-B2026.03 | 83.3 | |
| SigLIPClassifier=LDA, Embedding Projection=PCA<-, Backbone=ViT-B2026.03 | 83.3 | |
| SigLIP2Classifier=Prototype, Embedding Projection=PCA<-, Backbone=ViT-B2026.03 | 82.6 | |
| SigLIPClassifier=Prototype, Embedding Projection=PCA<-, Backbone=ViT-B2026.03 | 81.8 | |
| SigLIP2Classifier=Prototype, Embedding Projection=Original, Backbone=ViT-B2026.03 | 80.8 | |
| 2SFSBackbone=ViT-B/162025.03 | 80.7 | |
| SigLIPClassifier=Prototype, Embedding Projection=Original, Backbone=ViT-B2026.03 | 80.5 | |
| DINOv2Classifier=LDA, Embedding Projection=Original, Backbone=ViT-B2026.03 | 80.3 | |
| 2SFSBackbone=ViT-B/322025.03 | 76.8 | |
| CLIPClassifier=LDA, Embedding Projection=PCA<-, Backbone=ViT-B2026.03 | 76.5 | |
| CLIPClassifier=LDA, Embedding Projection=Original, Backbone=ViT-B2026.03 | 76.1 | |
| DINOv2Classifier=Prototype, Embedding Projection=Original, Backbone=ViT-B2026.03 | 76 | |
| CLIPClassifier=Prototype, Embedding Projection=PCA<-, Backbone=ViT-B2026.03 | 74.7 | |
| CLIPClassifier=Prototype, Embedding Projection=Original, Backbone=ViT-B2026.03 | 69.6 |