https://medium.com/analytics-vidhya/cnns-architectures-lenet-alexnet-vgg-googlenet-resnet-and-more-666091488df5
LeNet-5 (1998)
LeNet-5, a pioneering 7-level convolutional network by LeCun et al in 1998。
32x32 pixel greyscale inputimages。
AlexNet (2012)
It consisted 11x11, 5x5,3x3, convolutions, max pooling, dropout, data augmentation, ReLU activations, SGD with momentum.
It attached ReLU activations after every convolutional and fully-connected layer.
ZFNet(2013)
GoogLeNet/Inception(2014)
This module is based on several very small convolutions in order to drastically reduce the number of parameters.
Their architecture consisted of a 22 layer deep CNN but reduced the number of parameters from 60 million (AlexNet) to 4 million.
VGGNet (2014)
VGGNet consists of 16 convolutional layers and is very appealing because of its very uniform architecture.
only 3x3 convolutions, but lots of filters. Trained on 4 GPUs for 2–3 weeks. ResNet(2015)
ResNet(2015)
“skip connections” and features heavy batch normalization.
Such skip connections are also known as gated units or gated recurrent units and have a strong similarity to recent successful elements applied in RNNs.
152 layers while still having lower complexity than VGGNet.
AlexNet has parallel two CNN line trained on two GPUs with cross-connections, GoogleNet has inception modules, ResNet has residual connections.