January 2018
Beginner to intermediate
284 pages
8h 35m
English
As introduced in previous chapters, a typical CNN architecture consists of a series of layers, each of which transforms an input image tensor to an output tensor. Each of these layers may belong to more than one class. Each class of layers has a specific purpose in the network. The Sample ConvNet architecture figure shows an example of such a network, which is composed of input layer, convolutional layer, pooling layer, and fully connected layer (FC). A typical ConvNet can have an architecture such as [INPUT -> CONV -> POOL -> FC]. In this section, we will describe each of these layers in more detail and go over their role and significance for image processing:
Read now
Unlock full access