Image Classification on CIFAR-10 (CA, BA, ASR)
84.15Balanced AccuracyDSBA
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| DSBAPre-training Dataset=STL-102026.03 | 84.15 | — | 99.67 | |
| GhostEncoderPre-training Dataset=STL-102026.03 | 81.56 | — | 79.14 | |
| WaNetPre-training Dataset=STL-102026.03 | 81.42 | — | 11.48 | |
| IMPERATIVEPre-training Dataset=STL-102026.03 | 81.35 | — | 76.74 | |
| CTRLPre-training Dataset=STL-102026.03 | 75.19 | — | 62.33 | |
| BenignPre-training Dataset=STL-102026.03 | — | 83.22 | — |