I am preparing a different series in reinforcement learning. So I am taking a break in writing GAN for now. It is very hard to answer that topic quickly. Fortunately, this article should give you some good information.

Quote from the original paper:

“Our spectral normalization allows the parameter matrix to use as many features as possible while satisfying local 1-Lipschitz constraint”

So it wants to do what WGAN and WGAN-GP wants to achieve. But they doing it with spectra norm in which they also describe a easier way to compute it.

Written by

Deep Learning

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store