Class-Incremental Learning on Tiny-ImageNet 20 incremental tasks (test)
0.4863AccuracyDER
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| DERBase Technique=DER, Calibration=Vanilla2025.03 | 0.4863 | 0.4244 | 0.4243 | |
| DER+T-CILBase Technique=DER, Calibration=T-CIL2025.03 | 0.4833 | 0.059 | 0.0591 | |
| WA+T-CILBase Technique=WA, Calibration=T-CIL2025.03 | 0.368 | 0.1956 | 0.1957 | |
| WABase Technique=WA, Calibration=Vanilla2025.03 | 0.3599 | 0.1006 | 0.1006 | |
| EEIL+T-CILBase Technique=EEIL, Calibration=T-CIL2025.03 | 0.3422 | 0.2445 | 0.2446 | |
| EEILBase Technique=EEIL, Calibration=Vanilla2025.03 | 0.3354 | 0.2889 | 0.2889 | |
| ER+T-CILBase Technique=ER, Calibration=T-CIL2025.03 | 0.3062 | 0.1072 | 0.1079 | |
| ERBase Technique=ER, Calibration=Vanilla2025.03 | 0.3043 | 0.349 | 0.3488 |