While there are components common to many implementations of convolutional neural networks (CNN), there are some standard CNN architectures that have set high standards for industry performance. Alexnet first came to the forefront when it performed well on the ImageNet dataset. VGG was proposed a few years later to establish a more standardized implementation of CNNs.