AI Dictionary
Generative Adversarial Network (GAN)
Definition
A class of machine learning frameworks where two neural networks contest with each other in a game.
Deep Dive
A Generative Adversarial Network (GAN) is a class of machine learning frameworks composed of two neural networks, a 'generator' and a 'discriminator,' that are trained simultaneously in a zero-sum game. The generator's role is to create new data instances that resemble the real data distribution, while the discriminator's role is to distinguish between real data samples and the fake data produced by the generator. This adversarial process drives both networks to improve iteratively.
Examples & Use Cases
- 1Creating highly realistic synthetic human faces that are virtually indistinguishable from real photographs, used in art or virtual avatar creation
- 2Generating synthetic medical images (e.g., X-rays, MRI scans) to augment limited real datasets for training diagnostic AI models
- 3Performing image-to-image translation, such as transforming a rough sketch into a photorealistic image or changing a summer scene into a winter scene
Related Terms
Neural NetworkDeep LearningGenerative AI