A method for escaping limit cycles in training GANs

Li Keke, Yang Xinmin

This paper mainly conducts further research to alleviate the issue of limit cycling behavior in training generative adversarial networks (GANs) through the proposed predictive cent…