Face generation using a Deep Convolutional GAN

So far, we have seen how to generate new images. In this section, we will learn how to generate a new set of faces from an existing dataset of faces.

Getting ready

The approach we will be adopting for this exercise will be very similar to what we adopted in the Generating images using a Deep Convolutional GAN recipe:

1. Collect a dataset that contains multiple face images.
2. Generate random images at the start.

3. Train a discriminator by showing it a combination of faces and random images, where the discriminator is expected to differentiate between an actual face image and a generated face image.
4. Once the discriminator model is trained, freeze it and adjust the random images in such...