Generative Models
# LOGAN

Introduced by Wu et al. in LOGAN: Latent Optimisation for Generative Adversarial Networks
#### Papers

#### Tasks

#### Usage Over Time

#### Components

####
Categories

**LOGAN** is a generative adversarial network that uses a latent optimization approach using natural gradient descent (NGD). For the Fisher matrix in NGD, the authors use the empirical Fisher $F'$ with Tikhonov damping:

$$ F' = g \cdot g^{T} + \beta{I} $$

They also use Euclidian Norm regularization for the optimization step.

For LOGAN's base architecture, BigGAN-deep is used with a few modifications: increasing the size of the latent source from $186$ to $256$, to compensate the randomness of the source lost when optimising $z$. 2, using the uniform distribution $U\left(−1, 1\right)$ instead of the standard normal distribution $N\left(0, 1\right)$ for $p\left(z\right)$ to be consistent with the clipping operation, using leaky ReLU (with the slope of 0.2 for the negative part) instead of ReLU as the non-linearity for smoother gradient flow for $\frac{\delta{f}\left(z\right)}{\delta{z}}$ .

Source: LOGAN: Latent Optimisation for Generative Adversarial NetworksPaper | Code | Results | Date | Stars |
---|

Task | Papers | Share |
---|---|---|

Bias Detection | 2 | 20.00% |

Clustering | 2 | 20.00% |

Fairness | 1 | 10.00% |

Denoising | 1 | 10.00% |

BIG-bench Machine Learning | 1 | 10.00% |

Computational Efficiency | 1 | 10.00% |

Conditional Image Generation | 1 | 10.00% |

Image Generation | 1 | 10.00% |