EMPIRICAL ANALYSIS OF OVERFITTING AND MODE DROP IN GAN TRAINING

EMPIRICAL ANALYSIS OF OVERFITTING AND MODE DROP IN GAN TRAINING

IEEE International Conference on Image Processing

10.1109/ICIP40778.2020.9191083

https://ieeexplore.ieee.org/document/9191083

Yasin Yazici, Chuan-Sheng Foo, Stefan Winkler, Kim-Hui Yap, Vijay Chandrasekhar

generative adversarial networks, generative models, deep learning

26 October 2020

We examine two key questions in GAN training, namely overfitting and mode drop, from an empirical perspective. We show that when stochasticity is removed from the training procedure, GANs can overfit and exhibit almost no mode drop. Our results shed light on important characteristics of the GAN training procedure. They also provide evidence against prevailing intuitions that GANs do not memorize the training set, and that mode dropping is mainly due to properties of the GAN objective rather than how it is optimized during training.

http://creativecommons.org/licenses/by-nc-nd/4.0/

The computational work for this article was performed on resources of the National Supercomputing Centre, Singapore (https://www.nscc.sg). There is no specific funding for this work.

“© 2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.”

https://oar.a-star.edu.sg/communities-collections/articles/16574

Institute for Infocomm Research