Differentially Private Generative Adversarial Networks with Model Inversion

TL;DR

This paper introduces a novel differentially private GAN training method called DPMI, which improves output quality and convergence by mapping data to latent space before applying DP-GAN training, outperforming standard approaches.

Contribution

The paper proposes DPMI, a new approach that enhances differentially private GAN training by using model inversion and latent space mapping, leading to better results.

Findings

Outperforms standard DP-GAN in quality metrics

Achieves better convergence properties

Maintains privacy guarantees while improving utility

Abstract

To protect sensitive data in training a Generative Adversarial Network (GAN), the standard approach is to use differentially private (DP) stochastic gradient descent method in which controlled noise is added to the gradients. The quality of the output synthetic samples can be adversely affected and the training of the network may not even converge in the presence of these noises. We propose Differentially Private Model Inversion (DPMI) method where the private data is first mapped to the latent space via a public generator, followed by a lower-dimensional DP-GAN with better convergent properties. Experimental results on standard datasets CIFAR10 and SVHN as well as on a facial landmark dataset for Autism screening show that our approach outperforms the standard DP-GAN method based on Inception Score, Fr\'echet Inception Distance, and classification accuracy under the same privacy guarantee.