ML//GAN

A GAN (generative adversarial network) is a pair of networks playing a minimax game: the generator tries to create realistic samples, and the discriminator tries to tell real from fake.

A GAN (generative adversarial network) is a pair of networks playing a minimax game: the generator tries to create realistic samples, and the discriminator tries to tell real from fake.

Its training is notoriously unstable, with mode collapse, vanishing gradients and oscillation. GANs dominated image generation from 2014 to 2021, and then diffusion models took over.