Generative Adversarial Networks (GANs)
Generative Adversarial Networks, or GANs, consist of two neural networks, a generator and a discriminator, which compete against each other in a game-like scenario. The generator tries to create new data instances that resemble the real training data, while the discriminator tries to distinguish between these synthetic generations and real data. Through this adversarial process, both networks improve, with the generator eventually producing highly realistic, novel data.
It's like a counterfeit artist trying to make fake money so convincing that a detective, whose job is to spot fakes, struggles to tell it apart from real currency.
They enable the creation of highly realistic synthetic data, from images and videos to text and audio, with applications in art, data augmentation, and simulation.
Learn one new AI thing every day.
Daily Deck sends you seven plain-English cards like this every morning. Free.
Start free