Multi-domain Task Incremental Learning on MTIL Average 1.0 (test)
62Accuracy (Aircraft)Individual FT
Evaluation results
| Method | Links | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Individual FTFine-tuning Strategy=FT2025.03 | 62 | 95.1 | 89.6 | 79.5 | 98.9 | 97.5 | 92.7 | 99.6 | 94.7 | 89.6 | 81.8 | 89.2 | |
| ConDUFine-tuning Strategy=FT, Training Paradigm=Average2025.03 | 59.6 | 93.4 | 83.7 | 68.1 | 83.4 | 83.7 | 90.1 | 76.7 | 90.6 | 68.6 | 68.6 | 78.8 | |
| Dual-RAILTraining Paradigm=Average2025.03 | 52.5 | 96 | 80.6 | 70.4 | 81.3 | 86.3 | 89.1 | 73.9 | 90.2 | 68.5 | 66.5 | 77.8 | |
| MulKITraining Paradigm=Average2025.03 | 52.5 | 93.6 | 79.4 | 67 | 79.8 | 83.9 | 89.6 | 77.1 | 91.2 | 67.1 | 69.1 | 77.3 | |
| ConDUFine-tuning Strategy=LoRA, Training Paradigm=Average2025.03 | 51.9 | 94.9 | 84.4 | 69.8 | 81.1 | 84.4 | 90 | 77.3 | 89.5 | 69 | 69.3 | 78.3 | |
| DPeCLIPTraining Paradigm=Average2025.03 | 49.9 | 94.9 | 82.4 | 69.4 | 82.2 | 84.3 | 90 | 74 | 90.4 | 68.3 | 66.3 | 77.5 | |
| ZSCLTraining Paradigm=Average2025.03 | 45.1 | 92 | 80.1 | 64.3 | 79.5 | 81.6 | 89.6 | 75.2 | 88.9 | 64.7 | 68 | 75.4 | |
| Zero-shotFine-tuning Strategy=None2025.03 | 24.3 | 88.4 | 68.2 | 44.6 | 54.9 | 71 | 88.5 | 59.4 | 89 | 64.7 | 65.2 | 65.3 |