Table 3.
Classification accuracy using different transfer learning architectures for different epochs.
| Architectures | Epoch = 25 | Epoch = 50 | Epoch = 90 |
|---|---|---|---|
| AlexNet | 98.14 | 98.55 | 98.22 |
| GoogleNet | 95.69 | 97.16 | 97.24 |
| VGG-16 | 98.06 | 98.14 | 98.71 |
| VGG-19 | 97.97 | 98.55 | 98.47 |
| ResNet-18 | 96.01 | 97.86 | 97.81 |
| ResNet-50 | 96.67 | 97.65 | 96.16 |
| ResNet-101 | 96.67 | 96.83 | 95.99 |
| ResNet-inception-v2 | 93.67 | 95.03 | 95.50 |
| SENet | 56.66 | 56.66 | 95.18 |