Data-Incremental Learning on ImageNet-100 (5T)
76.67AccuracyMoCo
Evaluation results
| Method | Links | |
|---|---|---|
| MoCoStrategy=Joint2023.06 | 76.67 | |
| BarlowStrategy=Joint2023.06 | 75.89 | |
| BYOLStrategy=Joint2023.06 | 75.52 | |
| VICRegStrategy=Joint2023.06 | 75.08 | |
| SimCLRStrategy=Joint2023.06 | 71.91 | |
| BarlowStrategy=PNR2023.06 | 70.16 | |
| MoCoStrategy=PNR2023.06 | 69.98 | |
| BYOLStrategy=FT2023.06 | 69.76 | |
| BarlowStrategy=CaSSLe2023.06 | 69.24 | |
| VICRegStrategy=PNR2023.06 | 67.68 | |
| VICRegStrategy=CaSSLe2023.06 | 67.18 | |
| SimCLRStrategy=PNR2023.06 | 66.93 | |
| MoCoStrategy=CaSSLe2023.06 | 66.88 | |
| BarlowStrategy=FT2023.06 | 66.47 | |
| BYOLStrategy=CaSSLe2023.06 | 66.22 | |
| BYOLStrategy=PNR2023.06 | 66.08 | |
| SimCLRStrategy=CaSSLe2023.06 | 66.05 | |
| MoCoStrategy=FT2023.06 | 65.51 | |
| VICRegStrategy=FT2023.06 | 64.02 | |
| SimCLRStrategy=FT2023.06 | 62.88 |