In the previous chapters, we learned about predicting the class of an image and detecting where the object is located in the whole image. If we work backwards, we should be in a position to generate an image if we are given a class. Generative networks come in handy in this scenario, where we try to create new images that look very similar to the original image.
In this chapter, we will cover the following recipes:
- Generating images that can fool a neural network using an adversarial attack
- DeepDream algorithm to generate images
- Neural style transfer between images
- Generating images of digits using Generative Adversarial Networks
- Generating images of digits using a Deep Convolutional GAN
- Face generation using a Deep Convolutional GAN
- Face transition from one to another
- Performing vector arithmetic on generated images
