concept discriminator in category deep learning

This is an excerpt from Manning's book Deep Learning for Vision Systems MEAP V08 livebook. Login to get full access to this book.

As explained earlier, the goal of the discriminator is to predict if an image is real or fake. This is a typical supervised classification problem, so we can use the traditional classifier network that we learned in the previous chapters of this book. The network consists of stacked convolutional layers, followed by a dense output layer with sigmoid activation function. We use sigmoid activation function because this is a binary classification problem and the goal of the network is to output prediction probabilities values that range between 0 and 1. Where, 0 means that the image generated by the generator is fake and 1 means that it is 100% real.

Figure 8.5: The discriminator for the GAN

The discriminator is a normal and well understood classification model. As you can see in Figure 8.5, training the discriminator is pretty straight forward. As a supervised classification problem, we feed the discriminator with labeled images: fake (or generated) and real images. The real images come from the training dataset and the fake images are the output of the generator model.

# Build and compile the discriminator
       discriminator = discriminator_model()
discriminator.compile(loss='binary_crossentropy',optimizer='adam', metrics=['accuracy'])

copy