MGMD-GAN: Generalization Improvement of Generative Adversarial Networks with Multiple Generator Multiple Discriminator Framework Against Membership Inference Attacks

TL;DR

This paper introduces MGMD-GAN, a novel multi-generator, multi-discriminator framework that enhances generalization in GANs and reduces vulnerability to membership inference attacks by learning mixture distributions of data partitions.

Contribution

The paper proposes MGMD-GAN, a new architecture with multiple generators and discriminators trained on data partitions to improve generalization and defend against membership inference attacks.

Findings

MGMD-GAN reduces the generalization gap.

The model is less vulnerable to membership inference attacks.

Experimental results show improved robustness compared to other GANs.

Abstract

Generative Adversarial Networks (GAN) are among the widely used Generative models in various applications. However, the original GAN architecture may memorize the distribution of the training data and, therefore, poses a threat to Membership Inference Attacks. In this work, we propose a new GAN framework that consists of Multiple Generators and Multiple Discriminators (MGMD-GAN). Disjoint partitions of the training data are used to train this model and it learns the mixture distribution of all the training data partitions. In this way, our proposed model reduces the generalization gap which makes our MGMD-GAN less vulnerable to Membership Inference Attacks. We provide an experimental analysis of our model and also a comparison with other GAN frameworks.