Image Classification on CIFAR-10 (test) (Accuracy, Calibration, and Uncertainty Metrics)
92.6AccuracySAM
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| SAMBackbone=ResNet-20, Time=70s2026.03 | 92.6 | 0.22 | 1.56 | 0.11 | 92.5 | |
| q-VSGDBackbone=ResNet-20, Number of Monte-Carlo Samples=5, q=0.0, Time=173s2026.03 | 92.6 | 0.25 | 3.06 | 0.11 | 92.5 | |
| q-VSGDBackbone=ResNet-20, Number of Monte-Carlo Samples=5, q=0.6, Time=174s2026.03 | 92.6 | 0.25 | 3.04 | 0.11 | 92.5 | |
| IVON (1-MC)Backbone=ResNet-20, Number of Monte-Carlo Samples=1, Time=43s2026.03 | 92.5 | 0.26 | 3.43 | 0.12 | 92.5 | |
| q-VSGDBackbone=ResNet-20, Number of Monte-Carlo Samples=5, q=0.2, Time=173s2026.03 | 92.5 | 0.25 | 3.13 | 0.11 | 92.6 | |
| q-VSGDBackbone=ResNet-20, Number of Monte-Carlo Samples=5, q=0.4, Time=174s2026.03 | 92.4 | 0.25 | 3.17 | 0.11 | 92.7 | |
| q-VSGDBackbone=ResNet-20, Number of Monte-Carlo Samples=5, q=0.8, Time=182s2026.03 | 92.4 | 0.25 | 3.17 | 0.11 | 92.7 | |
| VSGDBackbone=ResNet-20, Number of Monte-Carlo Samples=5, Time=173s2026.03 | 92.4 | 0.25 | 3.13 | 0.12 | 92.6 | |
| q-VSGDBackbone=ResNet-20, Number of Monte-Carlo Samples=1, q=0.6, Time=43s2026.03 | 92.2 | 0.25 | 2.93 | 0.12 | 92.3 | |
| q-VSGDBackbone=ResNet-20, Number of Monte-Carlo Samples=1, q=0.0, Time=44s2026.03 | 92.1 | 0.25 | 3.01 | 0.12 | 92.3 | |
| q-VSGDBackbone=ResNet-20, Number of Monte-Carlo Samples=1, q=0.2, Time=43s2026.03 | 92.1 | 0.26 | 2.99 | 0.12 | 92.3 | |
| q-VSGDBackbone=ResNet-20, Number of Monte-Carlo Samples=1, q=0.8, Time=43s2026.03 | 92.1 | 0.26 | 3.02 | 0.12 | 92.2 | |
| VSGDBackbone=ResNet-20, Number of Monte-Carlo Samples=1, Time=42s2026.03 | 92.1 | 0.25 | 2.79 | 0.12 | 92.4 | |
| SGDBackbone=ResNet-20, Time=37s2026.03 | 92 | 0.28 | 3.84 | 0.12 | 92.1 | |
| q-VSGDBackbone=ResNet-20, Number of Monte-Carlo Samples=1, q=0.4, Time=43s2026.03 | 92 | 0.26 | 3.1 | 0.12 | 92.3 |