To start implementing GANs, we need to. It is hard to the quality of examples produced by GANs. A lower loss value doesn't always represent better quality. Often, for images, the only way to determine the quality is by visually inspecting the generated examples. We can than determine whether the generated images are realistic enough, more or less like a simple Turing test. In the following recipe, we will introduce GANs by using the well-known MNIST dataset and the Keras framework.
- We start by importing the necessary libraries, as follows:
import numpy as np from keras.models import Sequential, Model from keras.layers import Input, Dense, Activation, Flatten, Reshape from keras.layers import Conv2D, Conv2DTranspose, UpSampling2D from keras.layers import LeakyReLU, Dropout from keras.layers import BatchNormalization from keras.optimizers import Adam from keras import initializers from keras.datasets import mnist import matplotlib.pyplot as plt
- By using Keras...