Digest and Test: Text-to-Image Synthesis with GANs

In this chapter, we learned how to implement the discriminator and generator described in the paper “Generative Adversarial Text to Image Synthesis.” We learned how to implement the baseline trained on the Oxford-102 Flowers dataset. We also learned how to improve the baseline by using the matching-aware discriminator. Finally, we learned how to perform multiple types of inference using random sampling and interpolation in $Z$ space, interpolation on the text-embedding space, and arithmetic on the text-embedding space.

Test your understanding

Put your knowledge to the test with a quiz designed to reinforce your understanding.