TABLE 8.
Performance metric of the RIMONE dataset using different CNN models.
CNN model | AUC | ACC | SEN | SP | PRE | FM | GM |
---|---|---|---|---|---|---|---|
ResNet50 | 0.9721 | 0.9253 | 0.9200 | 0.9294 | 0.9109 | 0.9154 | 0.9247 |
VGG19 | 0.9330 | 0.8747 | 0.8900 | 0.8627 | 0.8357 | 0.8620 | 0.8763 |
AlexNet | 0.9089 | 0.7604 | 0.4950 | 0.9686 | 0.9252 | 0.6450 | 0.6924 |
Dns201 | 0.9513 | 0.8945 | 0.8450 | 0.9333 | 0.9086 | 0.8756 | 0.8881 |
IncRes | 0.9509 | 0.8989 | 0.8600 | 0.9294 | 0.9053 | 0.8821 | 0.8940 |
Fusion | — | 0.9495 | 0.9447 | 0.9843 | 0.9540 | 0.9447 | 0.9493 |
Bold values are showing the model results after applying classifier fusion operation.