Image Classification on ImageNet (test) (Specific Accuracy Metrics)
82.29Accuracy (voc-prior)GIT2
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GIT2Shots=5-shot per class2022.05 | 82.29 | 80.06 | 80.57 | |
| GITShots=5-shot per class2022.05 | 80.95 | 79.79 | 80.15 | |
| FlamingoShots=5-shot per class2022.05 | 77.3 | — | — | |
| GIT2Shots=1-shot per class2022.05 | 74.93 | 67.01 | 69.07 | |
| GITShots=1-shot per class2022.05 | 72.45 | 64.54 | 66.76 | |
| FlamingoShots=1-shot per class2022.05 | 71.7 | — | — | |
| GITLShots=5-shot per class2022.05 | 69.91 | 68.35 | 68.85 | |
| GITBShots=5-shot per class2022.05 | 53.4 | 51.99 | 52.39 | |
| GITLShots=1-shot per class2022.05 | 53.06 | 43.24 | 45.75 | |
| GIT2Shots=Zero-shot2022.05 | 40.57 | 1.91 | 41.92 | |
| GITBShots=1-shot per class2022.05 | 35.41 | 29.8 | 30.97 | |
| GITShots=Zero-shot2022.05 | 33.48 | 1.93 | 40.88 | |
| GITLShots=Zero-shot2022.05 | 10.35 | 0 | 15.28 | |
| GITBShots=Zero-shot2022.05 | 9.34 | 0 | 11.49 |