RAPID: Retrieval Augmented Training of Differentially Private Diffusion Models

TL;DR

RAPID introduces a retrieval-augmented training method for differentially private diffusion models, significantly improving generative quality and efficiency while maintaining privacy guarantees by leveraging public data as a knowledge base.

Contribution

It proposes a novel retrieval-augmented training approach for DP diffusion models that enhances utility and reduces resource requirements compared to existing methods.

Findings

Outperforms state-of-the-art DP diffusion models in quality and efficiency

Reduces memory footprint and inference cost significantly

Maintains strong differential privacy guarantees

Abstract

Differentially private diffusion models (DPDMs) harness the remarkable generative capabilities of diffusion models while enforcing differential privacy (DP) for sensitive data. However, existing DPDM training approaches often suffer from significant utility loss, large memory footprint, and expensive inference cost, impeding their practical uses. To overcome such limitations, we present RAPID: Retrieval Augmented PrIvate Diffusion model, a novel approach that integrates retrieval augmented generation (RAG) into DPDM training. Specifically, RAPID leverages available public data to build a knowledge base of sample trajectories; when training the diffusion model on private data, RAPID computes the early sampling steps as queries, retrieves similar trajectories from the knowledge base as surrogates, and focuses on training the later sampling steps in a differentially private manner. Extensive evaluation using benchmark datasets and models demonstrates that, with the same privacy guarantee, RAPID significantly outperforms state-of-the-art approaches by large margins in generative quality, memory footprint, and inference cost, suggesting that retrieval-augmented DP training represents a promising direction for developing future privacy-preserving generative models. The code is available at: https://github.com/TanqiuJiang/RAPID