参照：
[stanford]Convolutional Neural Networks (CNNs / ConvNets)
[karpathy]ConvNetJS CIFAR-10 demo
【知乎】没有博士学位，照样玩转TensorFlow深度学习
【视频】TensorFlow and deep learning, without a PhD
Architecture Overview
why ConvNet ?
Regular Neural Nets don’t scale well to full images. In CIFAR-10, images are only of size 32x32x3 (32 wide, 32 high, 3 color channels), so a single fully-connected neuron in a first hidden
layer of a regular Neural Network would have 32*32*3 = 3072 weights.