Table 9.
Training time and weights comparison of all models in the main dataset.
| Pre-Trained Model | Accuracy | Training Time (Second/Epoch) | Parameters (MB) |
|---|---|---|---|
| DenseNet201 | 0.9347 | 30 | 84.37 |
| VGG16 | 0.8857 | 26 | 63.08 |
| InceptionV3 | 0.8776 | 26 | 90.01 |
| ResNet50 | 0.7265 | 27 | 103.99 |
| ResNet50V2 | 0.9347 | 26 | 103.9 |
| ResNet152V2 | 0.9265 | 28 | 237.5 |
| Xception | 0.8898 | 27 | 93.48 |
| VGG19 | 0.8837 | 26 | 79.87 |
| ResNet101 | 0.7306 | 28 | 177.19 |
| ResNet101V2 | 0.9327 | 29 | 177.08 |
| NASNet | 0.9061 | 32 | 25.26 |
| MobileNetV2 | 0.9265 | 27 | 17.52 |
| MobileNet | 0.9571 | 25 | 19.33 |
| MobileNetV3Small | 0.5000 | 28 | 13.26 |
| InceptionResNetV2 | 0.8959 | 30 | 213.77 |
| EfficientNetB7 | 0.5000 | 34 | 263.42 |
| F-EDNC (Ours) | 0.9755 | 31 | 377.2 |
| O-EDNC (Ours) | 0.9632 | 31 | 337.24 |
| FC-EDNC (Ours) | 0.9714 | 31 | 348 |
| CANet (Ours) | 0.9163 | 26 | 338.7 |