Image Classification on CIFAR-10 (test) (Label Budget Evaluation)
81.45Accuracy (10% Budget)PruneFuse
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| PruneFusep=0.5, Selection Metric=Entropy, Architecture=ResNet-1102026.03 | 81.45 | 88.9 | 92.13 | 93.49 | 94.07 | |
| PruneFusep=0.5, Selection Metric=Least Conf., Architecture=ResNet-1102026.03 | 81.24 | 88.7 | 92.02 | 93.32 | 94.07 | |
| PruneFusep=0.6, Selection Metric=Least Conf., Architecture=ResNet-1102026.03 | 81.12 | 88.33 | 91.57 | 93.25 | 93.9 | |
| Baseline ALSelection Metric=Least Conf., Architecture=ResNet-1102026.03 | 80.74 | 87.8 | 91.5 | 93.19 | 93.68 | |
| PruneFusep=0.5, Selection Metric=Greedy k, Architecture=ResNet-1102026.03 | 80.4 | 87.77 | 90.74 | 92.48 | 93.53 | |
| PruneFusep=0.8, Selection Metric=Least Conf., Architecture=ResNet-1102026.03 | 80.34 | 88 | 91.22 | 92.89 | 93.8 | |
| PruneFusep=0.7, Selection Metric=Entropy, Architecture=ResNet-1102026.03 | 80.16 | 87.78 | 91.21 | 92.99 | 93.81 | |
| PruneFusep=0.5, Selection Metric=Random, Architecture=ResNet-1102026.03 | 80.08 | 86.52 | 89.48 | 90.82 | 91.79 | |
| PruneFusep=0.6, Selection Metric=Entropy, Architecture=ResNet-1102026.03 | 80.02 | 88.49 | 91.51 | 93.03 | 93.94 | |
| Baseline ALSelection Metric=Random, Architecture=ResNet-1102026.03 | 79.99 | 85.63 | 88.07 | 90.4 | 91.42 | |
| PruneFusep=0.7, Selection Metric=Least Conf., Architecture=ResNet-1102026.03 | 79.93 | 88.04 | 91.51 | 92.9 | 93.82 | |
| Baseline ALSelection Metric=Entropy, Architecture=ResNet-1102026.03 | 79.81 | 88.46 | 91.3 | 92.83 | 93.47 | |
| PruneFusep=0.8, Selection Metric=Entropy, Architecture=ResNet-1102026.03 | 79.61 | 88.12 | 90.94 | 92.76 | 93.54 | |
| PruneFusep=0.6, Selection Metric=Greedy k, Architecture=ResNet-1102026.03 | 79.44 | 87.05 | 90.3 | 92.15 | 93.22 | |
| PruneFusep=0.7, Selection Metric=Random, Architecture=ResNet-1102026.03 | 79.41 | 86.14 | 88.86 | 90.35 | 91.35 | |
| PruneFusep=0.8, Selection Metric=Random, Architecture=ResNet-1102026.03 | 78.94 | 86.2 | 89.11 | 90.5 | 91.42 | |
| Baseline ALSelection Metric=Greedy k, Architecture=ResNet-1102026.03 | 78.69 | 87.46 | 90.72 | 92.55 | 93.44 | |
| PruneFusep=0.7, Selection Metric=Greedy k, Architecture=ResNet-1102026.03 | 78.58 | 86.37 | 89.7 | 91.71 | 92.97 | |
| PruneFusep=0.6, Selection Metric=Random, Architecture=ResNet-1102026.03 | 78.55 | 85.94 | 88.77 | 90.66 | 92.02 | |
| PruneFusep=0.8, Selection Metric=Greedy k, Architecture=ResNet-1102026.03 | 78.41 | 85.9 | 89.57 | 91.38 | 92.21 |