Fig. (5)
Accuracy/Loss of the learning train obtained with (
a
) AlexNet, (
b
) VGG16, (
c
) VGG19 and (
d
) ProgNet.