Generative adversarial network: Difference between revisions

Revision as of 13:30, 12 November 2019

GANs are generative adversarial networks. They were developed by Ian Goodfellow.
Goal: Learn to generate examples from the same distribution as your training set.

Basis Structure

GANs consist of a generator and a discriminator.

For iteration i
  For iteration j
    Update Discriminator
  Update Generator

Variations

Wasserstein GAN

Paper
Medium post
This new WGAN-GP loss function improves the stability of training.
Normally, the discriminator is trained with a cross-entropy with sigmoid loss function.
The WGAN proposes using Wasserstein distance which is implemented by removing the cross-entropy+sigmoid and clipping (clamp) the weights on the discriminator to a range \(\displaystyle [-c, c]\).
However, weight clipping leads to other issues which limit the critic.
Instead of clipping, WGAN-GP proposes gradient penalty to enforce 1-Lipschitz .

Applications

CycleGan

InfoGAN

SinGAN

Paper
Website
Github Official PyTorch Implementation
SinGAN: Learning a Generative Model from a Single Natural Image

@@ Line 16: / Line 16: @@
 [https://arxiv.org/pdf/1704.00028.pdf Paper]<br>
 [https://medium.com/@jonathan_hui/gan-wasserstein-gan-wgan-gp-6a1a2aa1b490 Medium post]<br>
-This new WGAN-GP loss function improves the stability of training.
+This new WGAN-GP loss function improves the stability of training.<br>
+Normally, the discriminator is trained with a cross-entropy with sigmoid loss function.<br>
+The WGAN proposes using Wasserstein distance which is implemented by removing the cross-entropy+sigmoid
+and clipping (clamp) the weights on the discriminator to a range <math>[-c, c]</math>.<br>
+However, weight clipping leads to other issues which limit the critic.<br>
+Instead of clipping, WGAN-GP proposes gradient penalty to enforce 1-Lipschitz .
 ==Applications==