Identity-Focused Inference and Extraction Attacks on Diffusion Models

TL;DR

This paper introduces a new identity inference framework for diffusion models, revealing privacy risks by accurately detecting individuals' presence in training data, surpassing existing methods in success rate and AUC-ROC.

Contribution

It presents a novel identity inference approach for diffusion models, enabling accountability for training data inclusion, with comprehensive evaluations demonstrating superior performance over baseline methods.

Findings

Achieves up to 89% attack success rate and 0.91 AUC-ROC in membership inference.

Attains 92% accuracy in identity inference on LDM models trained on LFW.

Reaches 91.6% accuracy in data extraction attack on DDPMs.

Abstract

The increasing reliance on diffusion models for generating synthetic images has amplified concerns about the unauthorized use of personal data, particularly facial images, in model training. In this paper, we introduce a novel identity inference framework to hold model owners accountable for including individuals' identities in their training data. Our approach moves beyond traditional membership inference attacks by focusing on identity-level inference, providing a new perspective on data privacy violations. Through comprehensive evaluations on two facial image datasets, Labeled Faces in the Wild (LFW) and CelebA, our experiments demonstrate that the proposed membership inference attack surpasses baseline methods, achieving an attack success rate of up to 89% and an AUC-ROC of 0.91, while the identity inference attack attains 92% on LDM models trained on LFW, and the data extraction attack achieves 91.6% accuracy on DDPMs, validating the effectiveness of our approach across diffusion models.