Unseen prompt generalization on StanfordCars
74.89AccuracyCLIP
Evaluation Results
| Method | Links | |
|---|---|---|
| CLIPBackbone=ViT-B/16, Prompt=Fixed-prompt2022.10 | 74.89 | |
| Bayesian Prompt LearningBackbone=ViT-B/16, Prompt Length=4, Number of Shots=16, Monte Carlo samples (MC)=20, Epochs=402022.10 | 73.23 |