The generator creates candidate samples, while the discriminator learns to distinguish generated samples from real ones. Their opposing objectives provide the learning signal, but unstable dynamics and mode collapse can make training difficult.
A generative adversarial network trains a generator and a discriminator in competition to produce data resembling a training distribution.
The generator creates candidate samples, while the discriminator learns to distinguish generated samples from real ones. Their opposing objectives provide the learning signal, but unstable dynamics and mode collapse can make training difficult.