Generative Adversarial Network
Generative Adversarial Network — A family of generative models in which a generator and discriminator are trained with opposing objectives.
Technical Context
A family of generative models in which a generator and discriminator are trained with opposing objectives.
Training is sensitive to the balance between the two models; mode collapse, oscillation, and instability can limit diversity and reproducibility.
Related Articles
Direct source: The primary paper or official specification for Generative Adversarial Network is linked here for verification.