Image Classification on STL10 downstream
95.68Classification AccuracyNo Attack
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| No AttackAttack Type=None2024.05 | 95.68 | — | — | |
| BenignPre-train Dataset=CIFAR-10, Backbone=ResNet-18, SSL Algorithm=SimCLR2026.02 | 76.91 | — | — | |
| ADCAPre-train Dataset=CIFAR-10, Backbone=ResNet-18, SSL Algorithm=SimCLR2026.02 | 75.24 | — | 96.57 | |
| BADFSSPre-train Dataset=CIFAR-10, Backbone=ResNet-18, SSL Algorithm=SimCLR2026.02 | 74.51 | — | 72.46 | |
| BadEncoderPre-train Dataset=CIFAR-10, Backbone=ResNet-18, SSL Algorithm=SimCLR2026.02 | 74.29 | — | 64.37 | |
| DBAPre-train Dataset=CIFAR-10, Backbone=ResNet-18, SSL Algorithm=SimCLR2026.02 | 73.76 | — | 63.78 | |
| UBAPre-train Dataset=CIFAR-10, Backbone=ResNet-18, SSL Algorithm=SimCLR2026.02 | 72.37 | — | 70.85 | |
| FCBAPre-train Dataset=CIFAR-10, Backbone=ResNet-18, SSL Algorithm=SimCLR2026.02 | 71.43 | — | 66.32 | |
| INACTIVEAttack Type=Backdoor Attack2024.05 | — | 93.48 | 100 | |
| ISSBAAttack Type=Backdoor Attack2024.05 | — | 92.58 | 9.97 |