Qin, Yipeng ![]() ![]() |
Preview |
PDF
- Accepted Post-Print Version
Download (1MB) | Preview |
Preview |
PDF
- Supplemental Material
Download (62MB) | Preview |
Abstract
Despite the success of Lipschitz regularization in stabilizingGAN training, the exact reason of its effectiveness remains poorly un-derstood. The direct effect ofK-Lipschitz regularization is to restrict theL2-norm of the neural network gradient to be smaller than a thresholdK(e.g.,K= 1) such that‖∇f‖≤K. In this work, we uncover an evenmore important effect of Lipschitz regularization by examining its im-pact on the loss function:It degenerates GAN loss functions to almostlinear ones by restricting their domain and interval of attainable gradi-ent values. Our analysis shows that loss functions are only successful ifthey are degenerated to almost linear ones. We also show that loss func-tions perform poorly if they are not degenerated and that a wide rangeof functions can be used as loss function as long as they are sufficientlydegenerated by regularization. Basically, Lipschitz regularization ensuresthat all loss functionseffectively work in the same way.Empirically, weverify our proposition on the MNIST, CIFAR10 and CelebA datasets.
Item Type: | Conference or Workshop Item (Paper) |
---|---|
Date Type: | Publication |
Status: | Published |
Schools: | Computer Science & Informatics |
Publisher: | Springer |
ISBN: | 9783030585167 |
ISSN: | 0302-9743 |
Related URLs: | |
Date of First Compliant Deposit: | 28 July 2020 |
Date of Acceptance: | 2 July 2020 |
Last Modified: | 07 Nov 2022 10:51 |
URI: | https://orca.cardiff.ac.uk/id/eprint/133740 |
Citation Data
Cited 6 times in Scopus. View in Scopus. Powered By Scopus® Data
Actions (repository staff only)
![]() |
Edit Item |