Thyroid nodule diagnosis on TNCD (test)
73.28AccuracyAdaptive Curriculum Learning (ACL)
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Adaptive Curriculum Learning (ACL)Backbone=ResNet-182022.07 | 73.28 | 63.02 | 73.64 | 67.87 | 79.89 | |
| Adaptive Curriculum Learning (ACL)Backbone=DenseNet-1212022.07 | 72.92 | 60.74 | 77.8 | 67.99 | 79.84 | |
| Adaptive Curriculum Learning (ACL)Backbone=ConvNeXt-Tiny2022.07 | 72.5 | 61.19 | 74.75 | 67.21 | 78.85 | |
| Cross EntropyBackbone=ConvNeXt-Tiny2022.07 | 72.29 | 62.94 | 71.19 | 66.36 | 78.38 | |
| Curriculum Loss (CL)Backbone=ConvNeXt-Tiny2022.07 | 72.14 | 60.21 | 75.76 | 67 | 74.37 | |
| SuperLoss (SL)Backbone=DenseNet-1212022.07 | 71.82 | 64.38 | 66.86 | 65.42 | 78.98 | |
| SuperLoss (SL)Backbone=ConvNeXt-Tiny2022.07 | 71.47 | 62.66 | 68.39 | 65.36 | 77.87 | |
| SuperLoss (SL)Backbone=ResNet-182022.07 | 71.26 | 64.97 | 64.15 | 64.46 | 78.5 | |
| Cross EntropyBackbone=ResNet-182022.07 | 70.75 | 61.71 | 68.05 | 64.47 | 77.31 | |
| Curriculum Loss (CL)Backbone=ResNet-182022.07 | 70.54 | 62.87 | 65.85 | 64.04 | 76.98 | |
| Cross EntropyBackbone=DenseNet-1212022.07 | 70.36 | 63.92 | 63.31 | 63.31 | 78.05 | |
| Curriculum Loss (CL)Backbone=DenseNet-1212022.07 | 70.03 | 62.44 | 65.25 | 63.31 | 77.27 |