Are GANs Created Equal? A Large-Scale Study

by   Mario Lucic, et al.

Generative adversarial networks (GAN) are a powerful subclass of generative models. Despite a very rich research activity leading to numerous interesting GAN algorithms, it is still very hard to assess which algorithm(s) perform better than others. We conduct a neutral, multi-faceted large-scale empirical study on state-of-the art models and evaluation measures. We find that most models can reach similar scores with enough hyperparameter optimization and random restarts. This suggests that improvements can arise from a higher computational budget and tuning more than fundamental algorithmic changes. To overcome some limitations of the current metrics, we also propose several data sets on which precision and recall can be computed. Our experimental results suggest that future GAN research should be based on more systematic and objective evaluation procedures. Finally, we did not find evidence that any of the tested algorithms consistently outperforms the original one.


page 15

page 16


On Self Modulation for Generative Adversarial Networks

Training Generative Adversarial Networks (GANs) is notoriously challengi...

On Accurate Evaluation of GANs for Language Generation

Generative Adversarial Networks (GANs) are a promising approach to langu...

Activation Maximization Generative Adversarial Nets

Class label information has been empirically proven to be very useful in...

Wasserstein GAN Can Perform PCA

Generative Adversarial Networks (GANs) have become a powerful framework ...

An empirical study on evaluation metrics of generative adversarial networks

Evaluating generative adversarial networks (GANs) is inherently challeng...

Pros and Cons of GAN Evaluation Measures

Generative models, in particular generative adverserial networks (GANs),...

Pros and Cons of GAN Evaluation Measures: New Developments

This work is an update of a previous paper on the same topic published a...