Overview
GANs consist of a Generator that creates fake data and a Discriminator that tries to distinguish it from real data. Through competition, the generator becomes incredibly good at creating realistic outputs.
Use Cases
- Creating realistic human faces (Deepfakes).
- Style transfer (making a photo look like a painting).
- Data augmentation for training other models.
Challenges
GANs are notoriously difficult to train and can suffer from 'mode collapse,' where the generator only produces a limited variety of outputs.