Keras Deep Learning Cookbook

Introduction

Convolutional neural networks (CNNs) are networks of neurons that have learnable weights and biases. Every neuron accepts inputs, calculates a dot product, and follows it with a nonlinearity. CNNs are composed of several convolutional layers and are then followed by one or more fully connected layers, as in a standard multilayer neural network, starting from the raw image pixels on one end to class scores at the other. CNNs preserve the spatial relationship between pixels by learning feature representations. The feature is learned and applied across the whole image, allowing for the objects in the images to be shifted or translated in the scene and still be detectable by the network.

In a nutshell, CNNs are, fundamentally, several ...