Corpus ID: 49671009

The GAN Landscape: Losses, Architectures, Regularization, and Normalization

@article{Kurach2018TheGL,
  title={The GAN Landscape: Losses, Architectures, Regularization, and Normalization},
  author={Karol Kurach and Mario Lucic and Xiaohua Zhai and Marcin Michalski and S. Gelly},
  journal={ArXiv},
  year={2018},
  volume={abs/1807.04720}
}
Generative adversarial networks (GANs) are a class of deep generative models which aim to learn a target distribution in an unsupervised fashion. While they were successfully applied to many problems, training a GAN is a notoriously challenging task and requires a significant amount of hyperparameter tuning, neural architecture engineering, and a non-trivial amount of "tricks". The success in many practical applications coupled with the lack of a measure to quantify the failure modes of GANs… Expand
116 Citations
Regularization And Normalization For Generative Adversarial Networks: A Review
  • 1
  • PDF
Self-Supervised GANs via Auxiliary Rotation Loss
  • 119
  • PDF
Self-Supervised Generative Adversarial Networks
  • 17
  • PDF
Adversarial Latent Autoencoders
  • 41
  • PDF
On Self Modulation for Generative Adversarial Networks
  • 51
  • PDF
Diversity Regularized Adversarial Learning
  • 1
  • PDF
Diversity-Sensitive Conditional Generative Adversarial Networks
  • 72
  • PDF
PA-GAN: Improving GAN Training by Progressive Augmentation
  • 8
  • Highly Influenced
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 42 REFERENCES
Stabilizing Training of Generative Adversarial Networks through Regularization
  • 274
  • Highly Influential
  • PDF
Improved Training of Wasserstein GANs
  • 4,205
  • Highly Influential
  • PDF
Improved Techniques for Training GANs
  • 4,302
  • Highly Influential
  • PDF
On Self Modulation for Generative Adversarial Networks
  • 51
  • PDF
A Style-Based Generator Architecture for Generative Adversarial Networks
  • 1,883
  • PDF
Are GANs Created Equal? A Large-Scale Study
  • 538
  • PDF
Self-Attention Generative Adversarial Networks
  • 1,318
  • PDF
Large Scale GAN Training for High Fidelity Natural Image Synthesis
  • 1,694
  • PDF
...
1
2
3
4
5
...