Continual Learning on DomainNet (test)
76.84A5 AccuracyJoint Train
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Joint TrainBackbone=ViT-B/16, Pre-trained=ImageNet-21K, Training Strategy=Joint2026.02 | 76.84 | 81.25 | |
| EWC-LoRABackbone=ViT-B/16, Pre-trained=ImageNet-21K, Training Strategy=Sequential2026.02 | 73.46 | 79.58 | |
| SD-LoRABackbone=ViT-B/16, Pre-trained=ImageNet-21K, Training Strategy=Sequential2026.02 | 71.27 | 77.7 | |
| CL-LoRABackbone=ViT-B/16, Pre-trained=ImageNet-21K, Training Strategy=Sequential2026.02 | 71.06 | 77.76 | |
| InfLoRABackbone=ViT-B/16, Pre-trained=ImageNet-21K, Training Strategy=Sequential2026.02 | 71.01 | 77.75 | |
| CODA-PromptBackbone=ViT-B/16, Pre-trained=ImageNet-21K, Training Strategy=Sequential2026.02 | 70.58 | 76.68 | |
| L2PBackbone=ViT-B/16, Pre-trained=ImageNet-21K, Training Strategy=Sequential2026.02 | 70.26 | 75.83 | |
| Vanilla LoRABackbone=ViT-B/16, Pre-trained=ImageNet-21K, Training Strategy=Sequential2026.02 | 69.79 | 77.44 | |
| BiLoRABackbone=ViT-B/16, Pre-trained=ImageNet-21K, Training Strategy=Sequential2026.02 | 69.75 | 73.86 | |
| DualPromptBackbone=ViT-B/16, Pre-trained=ImageNet-21K, Training Strategy=Sequential2026.02 | 68.26 | 73.84 | |
| FinetuneBackbone=ViT-B/16, Pre-trained=ImageNet-21K, Training Strategy=Sequential2026.02 | 65.57 | 75.12 |