DP-RDM: Adapting Diffusion Models to Private Domains Without Fine-Tuning

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Lebensold, Jonathan, Sanjabi, Maziar, Astolfi, Pietro, Romero-Soriano, Adriana, Chaudhuri, Kamalika, Rabbat, Mike, Guo, Chuan
Format: Preprint
Publié: 2024
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866914792806023168
author Lebensold, Jonathan
Sanjabi, Maziar
Astolfi, Pietro
Romero-Soriano, Adriana
Chaudhuri, Kamalika
Rabbat, Mike
Guo, Chuan
author_facet Lebensold, Jonathan
Sanjabi, Maziar
Astolfi, Pietro
Romero-Soriano, Adriana
Chaudhuri, Kamalika
Rabbat, Mike
Guo, Chuan
contents Text-to-image diffusion models have been shown to suffer from sample-level memorization, possibly reproducing near-perfect replica of images that they are trained on, which may be undesirable. To remedy this issue, we develop the first differentially private (DP) retrieval-augmented generation algorithm that is capable of generating high-quality image samples while providing provable privacy guarantees. Specifically, we assume access to a text-to-image diffusion model trained on a small amount of public data, and design a DP retrieval mechanism to augment the text prompt with samples retrieved from a private retrieval dataset. Our \emph{differentially private retrieval-augmented diffusion model} (DP-RDM) requires no fine-tuning on the retrieval dataset to adapt to another domain, and can use state-of-the-art generative models to generate high-quality image samples while satisfying rigorous DP guarantees. For instance, when evaluated on MS-COCO, our DP-RDM can generate samples with a privacy budget of $ε=10$, while providing a $3.5$ point improvement in FID compared to public-only retrieval for up to $10,000$ queries.
format Preprint
id arxiv_https___arxiv_org_abs_2403_14421
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle DP-RDM: Adapting Diffusion Models to Private Domains Without Fine-Tuning
Lebensold, Jonathan
Sanjabi, Maziar
Astolfi, Pietro
Romero-Soriano, Adriana
Chaudhuri, Kamalika
Rabbat, Mike
Guo, Chuan
Machine Learning
Cryptography and Security
Computer Vision and Pattern Recognition
Text-to-image diffusion models have been shown to suffer from sample-level memorization, possibly reproducing near-perfect replica of images that they are trained on, which may be undesirable. To remedy this issue, we develop the first differentially private (DP) retrieval-augmented generation algorithm that is capable of generating high-quality image samples while providing provable privacy guarantees. Specifically, we assume access to a text-to-image diffusion model trained on a small amount of public data, and design a DP retrieval mechanism to augment the text prompt with samples retrieved from a private retrieval dataset. Our \emph{differentially private retrieval-augmented diffusion model} (DP-RDM) requires no fine-tuning on the retrieval dataset to adapt to another domain, and can use state-of-the-art generative models to generate high-quality image samples while satisfying rigorous DP guarantees. For instance, when evaluated on MS-COCO, our DP-RDM can generate samples with a privacy budget of $ε=10$, while providing a $3.5$ point improvement in FID compared to public-only retrieval for up to $10,000$ queries.
title DP-RDM: Adapting Diffusion Models to Private Domains Without Fine-Tuning
topic Machine Learning
Cryptography and Security
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2403.14421