Classification on MixedWM38 (test)
98.34AccuracyDenseNet121
Evaluation results
| Method | Links | ||
|---|---|---|---|
| DenseNet121Params=8M, Model Size (MB)=33.48, FLOPS=16.1M2022.07 | 98.34 | 0.9869 | |
| WSCNLoss Function=contrastive loss, Params=0.09M, Model Size (MB)=0.51, FLOPS=0.2M2022.07 | 98.2 | 0.9815 | |
| ResNet18Params=11.2M, Model Size (MB)=44.98, FLOPS=22.39M2022.07 | 97.96 | 0.9791 | |
| AlexNetParams=7.6M, Model Size (MB)=30.47, FLOPS=15.2M2022.07 | 97.91 | 0.9785 | |
| MobileNetV2Params=3.6M, Model Size (MB)=14.83, FLOPS=7.1M2022.07 | 97.91 | 0.9679 | |
| ResNet50Params=25.7M, Model Size (MB)=103.34, FLOPS=51.3M2022.07 | 97.19 | 0.9758 | |
| EfficientNetB0Params=5.4M, Model Size (MB)=22.14, FLOPS=10.7M2022.07 | 96.87 | 0.962 | |
| WSCNLoss Function=BCE loss, Params=0.08M, Model Size (MB)=0.51, FLOPS=0.2M2022.07 | 93.29 | 0.9622 | |
| DCNetParams=2.3M, Model Size (MB)=9.09, FLOPS=1.4M2022.07 | 93.2 | — | |
| LeNetParams=5.9M, Model Size (MB)=23.96, FLOPS=12.0M2022.07 | 76.44 | 0.7581 |