papers AI Learner
The Github is limit! Click to go to the new site.

Evaluating GANs via Duality

2018-11-13
Paulina Grnarova, Kfir Y Levy, Aurelien Lucchi, Nathanael Perraudin, Thomas Hofmann, Andreas Krause

Abstract

Generative Adversarial Networks (GANs) have shown great results in accurately modeling complex distributions, but their training is known to be difficult due to instabilities caused by a challenging minimax optimization problem. This is especially troublesome given the lack of an evaluation metric that can reliably detect non-convergent behaviors. We leverage the notion of duality gap from game theory in order to propose a novel convergence metric for GANs that has low computational cost. We verify the validity of the proposed metric for various test scenarios commonly used in the literature.

Abstract (translated by Google)
URL

https://arxiv.org/abs/1811.05512

PDF

https://arxiv.org/pdf/1811.05512


Similar Posts

Comments