Exemplar-Free Class-Incremental Learning on ImageNet
66.1Top-1 Incremental AccuracyFeTrIL
Evaluation results
| Method | Links | |
|---|---|---|
| FeTrILNumber of incremental steps (T)=52022.11 | 66.1 | |
| FeTrILNumber of incremental steps (T)=102022.11 | 65 | |
| DSLDANumber of incremental steps (T)=52022.11 | 64 | |
| DSLDANumber of incremental steps (T)=102022.11 | 63.8 | |
| FeTrILNumber of incremental steps (T)=202022.11 | 63.8 | |
| DSLDANumber of incremental steps (T)=202022.11 | 63.6 | |
| DeeSILNumber of incremental steps (T)=52022.11 | 61.9 | |
| DeeSILNumber of incremental steps (T)=102022.11 | 54.6 | |
| LUCIRNumber of incremental steps (T)=52022.11 | 47.4 | |
| DeeSILNumber of incremental steps (T)=202022.11 | 45.8 | |
| LUCIRNumber of incremental steps (T)=102022.11 | 37.2 | |
| LUCIRNumber of incremental steps (T)=202022.11 | 26.6 |