AI Glossary

Generative adversarial network

Also called GAN

What does “Generative adversarial network” mean?

Generative adversarial network: Proposed by Ian Goodfellow and colleagues in 2014, a GAN trains two networks against each other: a generator that creates fake samples and a discriminator that tries to tell fake from real. Competition pushes the generator toward increasingly realistic output.

Generative adversarial network visual guide: Proposed by Ian Goodfellow and colleagues in 2014, a GAN trains two networks against each other: a generator that creates fake samples and a discriminator that tries to tell fake from real. Competition pushes the generator toward increasingly realistic output.
Visual guide Open full size (new tab)

What is an example of “Generative adversarial network”?

GANs produced early photorealistic synthetic faces. Diffusion models have since overtaken them for most image generation, though GANs remain useful where speed matters.

Keep learning

See also

Diffusion modelTermVariational autoencoderTermLatent spaceTermDeepfakeTerm

Same topic

GPTTermFrontier modelTermEncoder-decoder modelTermLong short-term memoryTermEncoderTerm