Table 2 Performance and computational comparison of CNN models.
Pretrained CNN Models | Accuracy | Precision | Recall | F1 score | TT (s) | IT (s) | Model size (MB) |
|---|---|---|---|---|---|---|---|
DenseNet121 | 0.867 | 0.870 | 0.900 | 0.880 | 1200.980 | 14.830 | 34.500 |
MobileNet | 0.917 | 0.930 | 0.920 | 0.930 | 986.220 | 6.720 | 19.100 |
Xception | 0.806 | 0.830 | 0.840 | 0.810 | 1314.860 | 10.950 | 92.000 |
VGG16 | 0.946 | 0.940 | 0.950 | 0.950 | 1259.630 | 14.950 | 192.000 |
InceptionV3 | 0.940 | 0.950 | 0.950 | 0.950 | 1173.100 | 10.670 | 94.100 |
ResNet101V2 | 0.927 | 0.930 | 0.930 | 0.930 | 1326.300 | 8.770 | 173.000 |
