Accuracy on VTAB (Continual Learning)
95.5AccuracyJoint
Evaluation results
| Method | Links | |
|---|---|---|
| JointBackbone=ViT-B16, Pre-training=ImageNet21K, Evaluation Protocol=Offline, Initial Classes=2, New Classes per Task=12024.02 | 95.5 | |
| RanDumbBackbone=ViT-B16, Pre-training=ImageNet21K, Evaluation Protocol=Offline, Initial Classes=2, New Classes per Task=12024.02 | 92.4 | |
| RanPAC-impBackbone=ViT-B16, Pre-training=ImageNet21K, Evaluation Protocol=Offline, Initial Classes=2, New Classes per Task=12024.02 | 91.9 |