Class Incremental Learning on ImageNet-R Sequential (test)
82.41Last Top-1 AccuracyJoint
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| JointBackbone=ViT-B/16, Memory Buffer=None, Evaluation Protocol=CIL2025.03 | 82.41 | — | — | 93.27 | |
| VBM-SLCABackbone=ViT-B/16, Memory Buffer=None, Evaluation Protocol=CIL2025.03 | 80.07 | 84.61 | 94.86 | 92.74 | |
| SLCABackbone=ViT-B/16, Memory Buffer=None, Evaluation Protocol=CIL2025.03 | 78.81 | 83.58 | 94.39 | 92.02 | |
| CODA-PromptBackbone=ViT-B/16, Memory Buffer=None, Evaluation Protocol=CIL2025.03 | 76.79 | 81.89 | 93.3 | 90.49 | |
| L2PBackbone=ViT-B/16, Memory Buffer=None, Evaluation Protocol=CIL2025.03 | 73.56 | 76.76 | 89.39 | 86.61 | |
| DualPromptBackbone=ViT-B/16, Memory Buffer=None, Evaluation Protocol=CIL2025.03 | 69.12 | 73.31 | 89.19 | 85.46 | |
| S-iPromptBackbone=ViT-B/16, Memory Buffer=None, Evaluation Protocol=CIL2025.03 | 68.77 | 71.47 | 88.44 | 84.91 |