In this video, we discuss VGGNet, which had 16 convolution layers and hence the name VGG-16. It introduced simplicity and some kind of regularity in preparing CNN models.