/WGANGP-Presentation

Materials for my presentation on 17/07/2019 in Hong Kong Machine Learning Meetup. The topic is WGAN-GP (Wasserstein Generative Adversarial Network with Gradient Penalty)

Primary LanguageJupyter Notebook

WGAN-GP Presentation

Materials for my presentation on 17/07/2019 in Hong Kong Machine Learning Meetup. The topic is WGAN-GP (Wasserstein Generative Adversarial Network with Gradient Penalty)

Presentation Slide

You can view my presentation slide here

Gallery

image image

Dependencies

  1. tensorflow==1.5.0
  2. numpy==1.14.3
  3. pandas==0.23.4
  4. sklearn==0.20.3
  5. matplotlib==3.0.2
  6. scipy==1.3.0
  7. PIL==5.3.0
  8. imageio==2.4.1

Acknowledgement

Part of the code is borrowed from the following

  1. vanilla_gan
  2. improved_wasserstein_gan
  3. basic-gans

Reference

  1. Improved Training of Wasserstein GANs
  2. Wasserstein GAN
  3. Generative Adversarial Nets