GAN (Generative Adversarial Network)

Models and Architectures

Two neural networks locked in a contest - one generates fakes, the other judges them - until the fakes become convincing.

A GAN pits two networks against each other. The generator produces synthetic images and the discriminator tries to tell them from real ones. Each round the generator gets better at fooling the judge, and the judge gets sharper - until the generated output looks real.GANs produced the first photorealistic fake faces and kicked off the deepfake era. Diffusion models have since taken over most image generation, but GANs remain in use for photo enhancement, upscaling, and creating synthetic training data.