Designing GANs: A Likelihood Ratio Approach

TL;DR

This paper introduces a likelihood ratio metric for monitoring GAN training, proposes a simple design methodology for consistent adversarial problems, and demonstrates its effectiveness through experiments on standard datasets.

Contribution

It presents a novel likelihood ratio metric for GAN training analysis and a straightforward method for designing stable, consistent adversarial training problems.

Findings

Likelihood ratio effectively monitors GAN convergence and stability.

The proposed design methodology produces consistent adversarial training problems.

Experimental results validate the approach on well-known datasets.

Abstract

We are interested in the design of generative networks. The training of these mathematical structures is mostly performed with the help of adversarial (min-max) optimization problems. We propose a simple methodology for constructing such problems assuring, at the same time, consistency of the corresponding solution. We give characteristic examples developed by our method, some of which can be recognized from other applications, and some are introduced here for the first time. We present a new metric, the likelihood ratio, that can be employed online to examine the convergence and stability during the training of different Generative Adversarial Networks (GANs). Finally, we compare various possibilities by applying them to well-known datasets using neural networks of different configurations and sizes.