The relativistic discriminator: a key element missing from standard GAN

ICLR 2019 Alexia Jolicoeur-Martineau

In standard generative adversarial network (SGAN), the discriminator estimates the probability that the input data is real. The generator is trained to increase the probability that fake data is real. We show that this property can be induced by using a relativistic discriminator which estimate the probability that the given real data is more realistic than a randomly sampled fake data.

Full paper

Evaluation


Task Dataset Model Metric name Metric value Global rank Compare
Image Generation CAT 256x256 RaSGAN FID 32.11 # 1
Image Generation CIFAR-10 RSGAN-GP FID 25.60 # 19