Generative Adversarial Networks (GANs) consist of two competing neural networks: a generator and a discriminator.
Generator: Takes random noise as input and generates synthetic images that resemble real ones.
Discriminator: Evaluates whether an image is real (from the dataset) or fake (generated by the model).
Both networks train simultaneously, improving each other through adversarial learning. The generator learns to produce increasingly realistic images, while the discriminator enhances its ability to distinguish real from generated images. GANs are used for image generation, style transfer, super-resolution, and deepfake creation.