Pith. sign in

REVIEW

Likelihood Estimation for Generative Adversarial Networks

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 1707.07530 v1 pith:6GVRAZMG submitted 2017-07-24 cs.LG cs.AI

Likelihood Estimation for Generative Adversarial Networks

classification cs.LG cs.AI
keywords imagesgeneratedembeddingslikelihoodqualitysimpleadversarialdefine
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original 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.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.