Generative Models

## Bidirectional GAN

Introduced by Donahue et al. in Adversarial Feature Learning

A BiGAN, or Bidirectional GAN, is a type of generative adversarial network where the generator not only maps latent samples to generated data, but also has an inverse mapping from data to the latent representation. The motivation is to make a type of GAN that can learn rich representations for us in applications like unsupervised learning.

In addition to the generator $G$ from the standard GAN framework, BiGAN includes an encoder $E$ which maps data $\mathbf{x}$ to latent representations $\mathbf{z}$. The BiGAN discriminator $D$ discriminates not only in data space ($\mathbf{x}$ versus $G\left(\mathbf{z}\right)$), but jointly in data and latent space (tuples $\left(\mathbf{x}, E\left(\mathbf{x}\right)\right)$ versus $\left(G\left(z\right), z\right)$), where the latent component is either an encoder output $E\left(\mathbf{x}\right)$ or a generator input $\mathbf{z}$.

#### Latest Papers

PAPER DATE
Combining GANs and AutoEncoders for Efficient Anomaly Detection
| Fabio CarraraGiuseppe AmatoLuca BrombinFabrizio FalchiClaudio Gennaro
2020-11-16
2020-06-15
| Xinwei ShenTong ZhangKani Chen
2020-02-21
A critical analysis of self-supervision, or what we can learn from a single image
| Yuki M. AsanoChristian RupprechtAndrea Vedaldi
2019-04-30
An Empirical Study of Generative Models with Encoders
Paul K. RubensteinYunpeng LiDominik Roblek
2018-12-19
Semi-supervised learning with Bidirectional GANs
Maciej ZamorskiMaciej Zięba
2018-11-28
The Information-Autoencoding Family: A Lagrangian Perspective on Latent Variable Generative Modeling
Shengjia ZhaoJiaming SongStefano Ermon
2018-01-01
Theoretical limitations of Encoder-Decoder GAN architectures
Sanjeev AroraAndrej RisteskiYi Zhang
2017-11-07
Mickaël ChenLudovic Denoyer
2016-11-07