A classification-based study of covariate shift in GAN distributions
Name
1711.00970.pdf
Description
Accepted version
Size
2.35 MB
Format
Unknown
Checksum (MD5)
1754465d12b1537143d2b42a1f50f0ed
Author(s)
Madry, Aleksander
Schmidt, Ludwig
Santurkar, Shibani
Date Issued
2018
Citation
Madry, Aleksander, Schmidt, Ludwig and Santurkar, Shibani. 2018. "A classification-based study of covariate shift in GAN distributions."
Version
Author's final manuscript
Abstract
© 35th International Conference on Machine Learning, ICML 2018.All Rights Reserved. A basic, and still largely unanswered, question in the context of Generative Adversarial Networks (GANs) is whether they are truly able to capture all the fundamental characteristics of the distributions they are trained on. In particular, evaluating the diversity of GAN distributions is challenging and existing methods provide only a partial understanding of this issue. In this paper, we develop quantitative and scalable tools for assessing the diversity of GAN distributions. Specifically, we take a classification-based perspective and view loss of diversity as a form of covariate shift introduced by GANs. We examine two specific forms of such shift: mode collapse and boundary distortion. In contrast to prior work, our methods need only minimal human supervision and can be readily applied to state-of-the-art GANs on large, canonical dataseis. Examining popular GANs using our tools indicates that these GANs have significant problems in reproducing the more distributional properties of their training dataset.
MIT Department
Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Terms of Use
Creative Commons Attribution-Noncommercial-Share Alike
Persistent DSpace Link
DOI of Published Version
http://proceedings.mlr.press/v80/santurkar18a.html
