In an earlier section, we showed how the fundamental building blocks of a CNN consist of a series of convolutional and pooling layers. In this section, we're going to build a basic CNN consisting of this repeating pattern, as shown in the following diagram:
This basic CNN consists of two repeated blocks of Convolution and Max Pooling, following by two Fully Connected layers. As discussed in a previous section, the convolution and max pooling layers are responsible for learning the spatial characteristics of the classes (for example, identifying the ears of cats), whereas the Fully Connected layers learn to make predictions ...
