Likelihood estimation for generative adversarial networks

Autoren Hamid Eghbal-zadeh
Gerhard Widmer
Editoren
Titel Likelihood estimation for generative adversarial networks
Typ sonst
Organisation Workshop paper at ICML 2017 - 34th International Conference on Machine Learning, Workshop on Implicit Models, Sydney, Australia, August 6-11, 2017
Monat July
Jahr 2017
Seiten https://arxiv.org/abs/1707.07530
SCCH ID# 17065
Abstract

We present a simple method for assessing the quality of generated images in Generative Adversarial Networks (GANs). The method can be applied in any kind of GAN without interfering with the learning procedure or affecting the learning objective. The central idea is to define a likelihood function that correlates with the quality of the generated images. In particular, we derive a Gaussian likelihood function from the distribution of the embeddings (hidden activations) of the real images in the discriminator, and based on this, define two simple measures of how likely it is that the embeddings of generated images are from the distribution of the embeddings of the real images. This yields a simple measure of fitness for generated images, for all varieties of GANs. Empirical results on CIFAR-10 demonstrate a strong correlation between the proposed measures and the perceived quality of the generated images.