Class-Incremental Learning on CUB (T=10, test)
90.3Avg AccuracyRanPAC
Evaluation results
| Method | Links | ||
|---|---|---|---|
| RanPACAlgorithm=12023.07 | 90.3 | 25 | |
| Joint linear probeJoint training=true2023.07 | 88.7 | — | |
| RanPAC (No Phase 1)Phase 1=false2023.07 | 88.7 | — | |
| ADaM2023.07 | 87.1 | — | |
| RanPAC (No RPs or Phase 1)RP layer=false, Phase 1=false2023.07 | 86.8 | — | |
| RanPAC (No RPs)RP layer=false2023.07 | 86.6 | — | |
| RanPAC (NCM with Phase 1)Evaluation=NCM, Phase 1=true2023.07 | 85.4 | 28 | |
| RanPAC (NCM only)Evaluation=NCM, Phase 1=false2023.07 | 85.1 | 14 | |
| CODA-Prompt2023.07 | 79.5 | — | |
| DualPrompt2023.07 | 68.5 | — | |
| L2P2023.07 | 65.2 | — |