DP-CGAN: Differentially Private Synthetic Data and Label Generation

TLDR

Generative Adversarial Networks generate synthetic data, but preserving the privacy of individuals in training remains a major challenge. This work introduces a Differentially Private Conditional GAN framework that aims to improve model performance while safeguarding training data privacy. DP‑CGAN produces synthetic data and labels, employing a novel clipping and perturbation strategy and a Renyi differential privacy accountant to monitor the privacy budget. Experiments on MNIST demonstrate that DP‑CGAN yields visually and empirically promising results even with a single‑digit epsilon privacy parameter.

Abstract

Generative Adversarial Networks (GANs) are one of the well-known models to generate synthetic data including images, especially for research communities that cannot use original sensitive datasets because they are not publicly accessible. One of the main challenges in this area is to preserve the privacy of individuals who participate in the training of the GAN models. To address this challenge, we introduce a Differentially Private Conditional GAN (DP-CGAN) training framework based on a new clipping and perturbation strategy, which improves the performance of the model while preserving privacy of the training dataset. DP-CGAN generates both synthetic data and corresponding labels and leverages the recently introduced Renyi differential privacy accountant to track the spent privacy budget. The experimental results show that DP-CGAN can generate visually and empirically promising results on the MNIST dataset with a single-digit epsilon parameter in differential privacy.