Showing 1–1 of 1 results for author: Xinmin, Y

A method for escaping limit cycles in training GANs

Authors: Li Keke, Yang Xinmin

Abstract: This paper mainly conducts further research to alleviate the issue of limit cycling behavior in training generative adversarial networks (GANs) through the proposed predictive centripetal acceleration algorithm (PCAA). Specifically, we first derive the upper and lower bounds on the last-iterate convergence rates of PCAA for the general bilinear game, with the upper bound notably improving upon pre… ▽ More This paper mainly conducts further research to alleviate the issue of limit cycling behavior in training generative adversarial networks (GANs) through the proposed predictive centripetal acceleration algorithm (PCAA). Specifically, we first derive the upper and lower bounds on the last-iterate convergence rates of PCAA for the general bilinear game, with the upper bound notably improving upon previous results. Then, we combine PCAA with the adaptive moment estimation algorithm (Adam) to propose PCAA-Adam, a practical approach for training GANs. Finally, we validate the effectiveness of the proposed algorithm through experiments conducted on bilinear games, multivariate Gaussian distributions, and the CelebA dataset, respectively. △ Less

Submitted 11 August, 2023; v1 submitted 7 October, 2020; originally announced October 2020.