Figure 4
The image contains four line graphs labeled VGG-16, ResNet-50, MobileNet, and Inception. Each graph displays training accuracy in red and validation accuracy in blue over 100 epochs. The x-axis represents the number of epochs, while the y-axis represents accuracy. In the VGG-16 graph, both training and validation accuracy gradually increase, with training accuracy consistently higher. The ResNet-50 graph shows a similar trend with both accuracies rising steadily. The MobileNet graph indicates a rapid increase in training accuracy early on, which then stabilizes, while validation accuracy fluctuates but generally increases. The Inception graph also shows a steady rise in both training and validation accuracy, with training accuracy slightly higher. All values are approximated.Training and validation accuracy curves for the four evaluated architectures: (a) VGG-16, (b) ResNet-50V2, (c) MobileNetV2 and (d) InceptionV3