7 Generative Adversarial Networks


  • The basic GAN structure and relationship to game-theory. Basic impleminintation. Usage of GANs: Deep fakes, data-augmentation, other uses.

  • Generative modelling is an unsupervised learning task in machine learning that involves automatically discovering and learning the regularities or patterns in input data in such a way that the model can be used to generate or output new examples that plausibly could have been drawn from the original dataset.PLAG

  • Generative Adversarial Networks (GANs)

  • GANs are useful for semi-supervised learning, unsupervised learning, and most notably for generation.

  • Latent space

7.1 Generative Models

7.2 The idea of Generative Adversarial Networks (GAN)

\[ \min _{G} \max _{D} V(D, G)=\mathbb{E}_{\boldsymbol{x} \sim p_{\text {data }}(\boldsymbol{x})}[\log D(\boldsymbol{x})]+\mathbb{E}_{\boldsymbol{z} \sim p_{\boldsymbol{z}}(\boldsymbol{z})}[\log (1-D(G(\boldsymbol{z})))] \]

7.3 The basic GAN details

7.4 Quality Measures

  • Inception Score
  • Fr├ęchet Inception Distance (FID)

7.5 Modification of GANs

7.6 Style Transfer

7.7 Applications

See the GAN ZOO.

they are more resilient to vanishing gradients

Page built: 2021-01-17