Image Classification on ImageNet-100 20 tasks 5 classes per task (test)
86.26AccuracyHydraCIL
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| HydraCILk (number of prototypes)=202026.06 | 86.26 | 7.29 | 0.021 | 0.004 | |
| HydraCILk (number of prototypes)=22026.06 | 85.72 | 7.32 | 0.021 | 0.004 | |
| HydraCILk (number of prototypes)=52026.06 | 85.48 | 6.94 | 0.02 | 0.003 | |
| CIFNet2026.06 | 78.1 | 71.5 | 0.271 | 0.047 | |
| DER2026.06 | 63.66 | 1,354.4 | 9.972 | 1.736 | |
| RMM-FOSTER2026.06 | 59.46 | 935.67 | 6.255 | 1.089 | |
| Fine-tune2026.06 | 4.64 | 188.93 | 1.136 | 0.198 |