Image Classification on CIFAR100 10-split
87.94AccuracyFedProTIP
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| FedProTIPtask identity prediction=with, Data Distribution Scenario=IID, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 87.94 | 0.34 | |
| FedProTIPtask identity prediction=with, Data Distribution Scenario=alpha = 0.5, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 86 | 0.83 | |
| FedProTIPtask identity prediction=with, Data Distribution Scenario=alpha = 0.2, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 81.94 | 1.35 | |
| FedProTIP (-t)task identity prediction=without, Data Distribution Scenario=IID, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 52.3 | 15.66 | |
| FedProTIP (-t)task identity prediction=without, Data Distribution Scenario=alpha = 0.5, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 48.41 | 15.59 | |
| FOTData Distribution Scenario=IID, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 46.86 | 21.11 | |
| FedProTIP (-t)task identity prediction=without, Data Distribution Scenario=alpha = 0.2, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 42.19 | 14.91 | |
| FOTData Distribution Scenario=alpha = 0.5, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 41.8 | 20.86 | |
| LANDERData Distribution Scenario=IID, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 39.09 | 9.27 | |
| LANDERData Distribution Scenario=alpha = 0.5, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 37.59 | 10.21 | |
| FOTData Distribution Scenario=alpha = 0.2, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 34.65 | 18.09 | |
| TARGETData Distribution Scenario=IID, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 29.56 | 42.73 | |
| TARGETData Distribution Scenario=alpha = 0.5, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 27.37 | 37.6 | |
| LANDERData Distribution Scenario=alpha = 0.2, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 23.56 | 13.28 | |
| TARGETData Distribution Scenario=alpha = 0.2, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 23.05 | 34.63 | |
| FedAvgData Distribution Scenario=IID, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 18.92 | 63.2 | |
| FedAvgData Distribution Scenario=alpha = 0.2, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 15.76 | 52.8 | |
| FedAvgData Distribution Scenario=alpha = 0.5, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 15.35 | 62.9 | |
| LGAData Distribution Scenario=IID, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 14.93 | 72.06 | |
| LGAData Distribution Scenario=alpha = 0.5, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 14.35 | 71.09 | |
| GLFCData Distribution Scenario=IID, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 14.07 | 69.17 | |
| GLFCData Distribution Scenario=alpha = 0.5, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 11.86 | 68.2 | |
| LGAData Distribution Scenario=alpha = 0.2, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 11.67 | 65.82 | |
| GLFCData Distribution Scenario=alpha = 0.2, Backbone=ResNet-18, Pre-training=ImageNet-1K, Local Epochs=5, Global rounds per task=502025.09 | 10.33 | 63.98 |