Class-Incremental Learning on Stanford-Cars (5 tasks)
84.4AccuracyJoint
Evaluation results
| Method | Links | |
|---|---|---|
| JointBackbone=CLIP ViT-B/16, Method Category=Baseline2026.05 | 84.4 | |
| TRMBackbone=CLIP ViT-B/16, Method Category=Model Merging2026.05 | 73.2 | |
| CLG-CBMYear=2025, Backbone=CLIP ViT-B/16, Method Category=PEFT2026.05 | 72.8 | |
| DualPromptYear=2023, Backbone=CLIP ViT-B/16, Method Category=PEFT2026.05 | 68.9 | |
| MagMaxYear=2024, Backbone=CLIP ViT-B/16, Method Category=Model Merging2026.05 | 68.5 | |
| TIESYear=2023, Backbone=CLIP ViT-B/16, Method Category=Model Merging2026.05 | 67.9 | |
| L2PYear=2022, Backbone=CLIP ViT-B/16, Method Category=PEFT2026.05 | 67.8 | |
| CODAPromptYear=2023, Backbone=CLIP ViT-B/16, Method Category=PEFT2026.05 | 67.4 | |
| Model StockYear=2024, Backbone=CLIP ViT-B/16, Method Category=Model Merging2026.05 | 66.6 | |
| EWCYear=2017, Backbone=CLIP ViT-B/16, Method Category=Conventional2026.05 | 65.8 | |
| LwFYear=2017, Backbone=CLIP ViT-B/16, Method Category=Conventional2026.05 | 65.2 | |
| Zero-shotBackbone=CLIP ViT-B/16, Method Category=Baseline2026.05 | 64.4 |