Table 3.
Summary of performance metrics of Fet-Net and prominent CNN architectures during the validation experiment.
| Architecture | Accuracy (%) | Loss | Number of Parameters |
|---|---|---|---|
| Fet-Net (Average of 3 Seeds) | 82.20 | 0.4777 | 10,556,420 |
| VGG16 | 63.80 | 1.6586 | 14,847,044 |
| VGG19 | 61.82 | 1.6588 | 20,156,740 |
| ResNet-50 | 53.06 | 1.7716 | 24,113,284 |
| ResNet-50V2 | 70.58 | 1.1676 | 24,090,372 |
| ResNet-101 | 57.85 | 1.3354 | 43,183,748 |
| ResNet-101V2 | 66.12 | 1.4789 | 43,152,132 |
| ResNet-152 | 60.00 | 1.2846 | 58,896,516 |
| ResNet-152V2 | 76.86 | 1.0471 | 58,857,220 |
| Inception-ResNetV2 | 63.64 | 1.5332 | 54,731,236 |
| InceptionV3 | 59.17 | 1.7725 | 22,328,356 |
| Xception | 62.48 | 1.5365 | 21,387,052 |