Image Classification on CIFAR-100 (test) (Accuracy and Calibration Profile)
83.07Accuracy (Test)CSAM
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| CSAMBackbone=ResNet-562025.05 | 83.07 | 1.93 | 0.15 | 1.99 | 1.54 | 96.07 | |
| SAMBackbone=ResNet-562025.05 | 82.93 | 2.11 | 0.17 | 2.17 | 1.89 | 94.15 | |
| MixupBackbone=ResNet-562025.05 | 82.09 | 4.28 | 0.18 | 4.24 | 4.2 | 94.35 | |
| BalCALBackbone=ResNet-562025.05 | 81.34 | 5.69 | 0.18 | 5.66 | 2.64 | 93.49 | |
| MIT-LBackbone=ResNet-562025.05 | 81.29 | 3.26 | 0.18 | 3.24 | 3.09 | 94.76 | |
| MMCEBackbone=ResNet-562025.05 | 81.02 | 4.02 | 0.18 | 3.96 | 3.69 | 93.83 | |
| CEBackbone=ResNet-562025.05 | 81.01 | 3.95 | 0.21 | 3.86 | 3.38 | 93.93 | |
| MIMOBackbone=ResNet-562025.05 | 80.75 | 2.38 | 0.17 | 2.31 | 2.04 | 95.14 | |
| DualFocalBackbone=ResNet-562025.05 | 80.74 | 2.68 | 0.18 | 2.66 | 2.24 | 94.81 | |
| AdaFocalBackbone=ResNet-562025.05 | 80.7 | 2.58 | 0.19 | 2.61 | 2.31 | 93.75 | |
| bSAMBackbone=ResNet-562025.05 | 80.59 | 8.27 | 0.22 | 8.27 | 2.59 | 94.01 | |
| Focal Loss (FL)Backbone=ResNet-562025.05 | 80.55 | 2.84 | 0.19 | 2.79 | 2.75 | 94.43 | |
| ACLSBackbone=ResNet-562025.05 | 80.49 | 6.38 | 0.19 | 6.31 | 2.9 | 93.16 | |
| Rank1-BNNBackbone=ResNet-562025.05 | 80.21 | 3.59 | 0.19 | 3.57 | 2.42 | 94.29 | |
| BatchEnsembleBackbone=ResNet-562025.05 | 79.93 | 6.86 | 0.21 | 6.77 | 2.49 | 94.15 | |
| VIBackbone=ResNet-562025.05 | 76.3 | 10.29 | 0.27 | 10.29 | 2.08 | 92.62 |