Table 7 Results summary of the comparative evaluation.
Model | Type | Accuracy (%) | Number of parameters (millions) | Computational efficiency | Inference time (seconds) |
|---|---|---|---|---|---|
VGG16 | CNN | 93.52 | 138 | Moderate | 10.35 |
ResNet50 | CNN | 97.63 | 25.6 | High | 7.61 |
DenseNet121 | CNN | 97.22 | 8 | High | 11.05 |
DenseNet169 | CNN | 98.63 | 14.3 | Very High | 15.76 |
DenseNet201 | CNN | 98.92 | 20.2 | Very High | 20.56 |
EfficientNet-B2 | CNN | 98.29 | 9.2 | Optimized | 9.46 |
ViT | Transformer | 95.3 | 86 | Low | 18.42 |
DeiT | 82.52 | 22 | High | 19.43 | |
Swin transformer | Transformer | 90.14 | 29 | Low | 22.67 |
