DP-CGAN: Differentially Private Synthetic Data and Label Generation

TL;DR

This paper introduces DP-CGAN, a differentially private conditional GAN framework that generates synthetic data and labels while effectively preserving privacy, demonstrated on MNIST with promising results under strict privacy constraints.

Contribution

The paper proposes a novel clipping and perturbation strategy for training DP-CGAN, improving synthetic data quality while maintaining differential privacy.

Findings

Generates high-quality synthetic data with strong privacy guarantees

Uses Renyi differential privacy accountant for accurate privacy tracking

Achieves promising results on MNIST with low epsilon values

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.