dp-promise: Differentially Private Diffusion Probabilistic Models for Image Synthesis
Haichen Wang, Shuchao Pang, Zhigang Lu, Yihang Rao, Yongbin Zhou, Minhui Xue
33rd USENIX Security Symposium · Day 1 · USENIX Security '24
This talk introduces **DP-promise**, a novel framework for training **differentially private diffusion probabilistic models** designed for image synthesis. Presented by Haichen Wang from Nanjing University of Science and Technology, alongside collaborators from James Cook University and CSIRO's Data61, the work addresses a critical challenge in the era of data-driven deep learning: the tension between the need for large datasets and the imperative to protect individual privacy. While deep learning models, particularly generative models like **Generative Adversarial Networks (GANs)** and **diffusion models**, thrive on vast amounts of data, using sensitive information—such as medical images or facial recognition data—raises significant privacy concerns. Even synthetic images generated by these models have been shown to inadvertently leak information from their original training data, necessitating robust privacy-preserving mechanisms.
AI review
This paper introduces DP-promise, a novel two-phase framework for differentially private image synthesis using diffusion models. It ingeniously leverages the inherent noise in the diffusion forward process to achieve privacy, significantly improving the privacy-utility trade-off compared to traditional DP-SGD, and demonstrating superior performance in generating high-quality synthetic images with strong theoretical guarantees.