BriefGPT.xyz
Jun, 2017
通过双时间尺度更新规则训练的GAN收敛于局部纳什均衡
GANs Trained by a Two Time-Scale Update Rule Converge to a Nash Equilibrium
HTML
PDF
Martin Heusel, Hubert Ramsauer, Thomas Unterthiner, Bernhard Nessler, Günter Klambauer...
TL;DR
本文提出一种基于二次尺度更新规则的GAN训练方法,使用随机逼近理论证明其可以收敛于本地Nash均衡点,并且在图像生成方面的表现优于传统GAN算法,同时提出了一种新的用于评估GAN生成图像质量的指标Fr\'echet Inception Distance。
Abstract
generative adversarial networks
(GANs) excel at creating realistic images with complex models for which maximum likelihood is infeasible. However, the convergence of
gan training
has still not been proved. We pro
→