Image Classification on CIFAR-10-C OOD corruptions
63.37AccuracyPruneFuse
Evaluation Results
| Method | Links | |
|---|---|---|
| PruneFuseLabel Budget (b)=50%, Backbone=ResNet-56, Pruning ratio (p)=0.52026.03 | 63.37 | |
| PruneFuseLabel Budget (b)=40%, Backbone=ResNet-56, Pruning ratio (p)=0.52026.03 | 58.36 | |
| Baseline (AL)Label Budget (b)=50%, Backbone=ResNet-562026.03 | 58.32 | |
| Baseline (AL)Label Budget (b)=40%, Backbone=ResNet-562026.03 | 57.27 | |
| PruneFuseLabel Budget (b)=30%, Backbone=ResNet-56, Pruning ratio (p)=0.52026.03 | 54.6 | |
| Baseline (AL)Label Budget (b)=30%, Backbone=ResNet-562026.03 | 53.62 | |
| PruneFuseLabel Budget (b)=20%, Backbone=ResNet-56, Pruning ratio (p)=0.52026.03 | 52.44 | |
| Baseline (AL)Label Budget (b)=20%, Backbone=ResNet-562026.03 | 48.56 | |
| PruneFuseLabel Budget (b)=10%, Backbone=ResNet-56, Pruning ratio (p)=0.52026.03 | 44.58 | |
| Baseline (AL)Label Budget (b)=10%, Backbone=ResNet-562026.03 | 42.95 |