Which 2014 paper introduced generative adversarial networks, a framework pairing a generator with a discriminator?

“Generative Adversarial Nets” was the 2014 paper that introduced generative adversarial networks.

The paper was written by Ian Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. It proposed training two neural networks in competition: a generator creates synthetic examples, while a discriminator tries to distinguish generated examples from real data.

This adversarial setup turned generation into a learned competition. As training progresses, the generator attempts to produce increasingly convincing outputs, while the discriminator becomes better at detecting fakes. GANs later became important in image synthesis, style transfer, super-resolution, and data augmentation. The paper is sometimes confused with work on diffusion models or variational autoencoders, which use different generation mechanisms. It is also distinct from the later Transformer paper, which addressed sequence modeling rather than adversarial generation.