A Variational Inequality Perspective on Generative Adversarial Networks
This work addresses the training instability problem for researchers and practitioners using GANs, offering incremental improvements through optimization techniques.
The authors tackled the difficulty of training Generative Adversarial Networks (GANs) by framing their optimization as a variational inequality problem, leading to improved training stability and convergence with methods like averaging and extrapolation.
Generative adversarial networks (GANs) form a generative modeling approach known for producing appealing samples, but they are notably difficult to train. One common way to tackle this issue has been to propose new formulations of the GAN objective. Yet, surprisingly few studies have looked at optimization methods designed for this adversarial training. In this work, we cast GAN optimization problems in the general variational inequality framework. Tapping into the mathematical programming literature, we counter some common misconceptions about the difficulties of saddle point optimization and propose to extend techniques designed for variational inequalities to the training of GANs. We apply averaging, extrapolation and a computationally cheaper variant that we call extrapolation from the past to the stochastic gradient method (SGD) and Adam.