mlsanityData to 7 Oct 2026

PAC Privacy Preserving Diffusion Models

arXiv 2312.01201 ↗First posted 2 Dec 2023cs.LG cs.AIcited by 0

On the map: Diffusion models → Robustness, privacy and adversarial aspects

Abstract (arXiv)

Data privacy protection is garnering increased attention among researchers. Diffusion models (DMs), particularly with strict differential privacy, can potentially produce images with both high privacy and visual quality. However, challenges arise such as in ensuring robust protection in privatizing specific data attributes, areas where current models often fall short. To address these challenges, we introduce the PAC Privacy Preserving Diffusion Model, a model leverages diffusion principles and ensure Probably Approximately Correct (PAC) privacy. We enhance privacy protection by integrating a private classifier guidance into the Langevin Sampling Process. Additionally, recognizing the gap in measuring the privacy of models, we have developed a novel metric to gauge privacy levels. Our model, assessed with this new metric and supported by Gaussian matrix computations for the PAC bound, has shown superior performance in privacy protection over existing leading private generative models according to benchmark tests.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Robustness, privacy and adversarial aspects, one of 14 directions in Diffusion models. Builds on 27 of its 37 references that are in the map. First posted on arXiv 2 Dec 2023.

Builds on

27 of 37 references.

Closest ancestors · 4

Same direction, newest first

  1. 2023DP-LDMs: Differentially Private Latent Diffusion Models25 May
  2. 2023Differentially Private Diffusion Models Generate Useful Synthetic Images27 Feb
  3. 2023Membership Inference Attacks against Diffusion Models7 Feb
  4. 2022Differentially Private Diffusion Models18 Oct

Other references · 23

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  2. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 8,964
  3. 2021Diffusion Models Beat GANs on Image Synthesiscited by 8,185
  4. 2022Classifier-Free Diffusion Guidancecited by 5,717
  5. 2015Rethinking the Inception Architecture for Computer Visioncited by 5,451
  6. 2019Generative Modeling by Estimating Gradients of the Data Distributioncited by 4,067
  7. 2014Deep Learning Face Attributes in the Wildcited by 3,679
  8. 2016Improved Techniques for Training GANscited by 2,938
  9. 2016Deep Learning with Differential Privacycited by 2,682
  10. 2020Improved Techniques for Training Score-Based Generative Modelscited by 1,036
  11. 2018Differentially Private Generative Adversarial Networkcited by 235
  12. 2021Conditional Image Generation with Score-Based Diffusion Modelscited by 195
11 more
  1. 2020DP-CGAN: Differentially Private Synthetic Data and Label Generationcited by 118
  2. 2020GS-WGAN: A Gradient-Sanitized Approach for Learning Differentially Private Generatorscited by 117
  3. 2020DP-MERF: Differentially Private Mean Embeddings with Random Features for Practical Privacy-Preserving Data Generationcited by 78
  4. 2021Don't Generate Me: Training Differentially Private Generative Models with Sinkhorn Divergencecited by 54
  5. 2023SoK: Privacy-Preserving Data Synthesiscited by 37
  6. 2016MOON: A Mixed Objective Optimization Network for the Recognition of Facial Attributescited by 34
  7. 2022Pre-trained Perceptual Features Improve Differentially Private Image Generationcited by 28
  8. 2020P3GM: Private High-Dimensional Data Release via Privacy Preserving Phased Generative Modelcited by 21
  9. 2022PAC Privacy: Automatic Privacy Measurement and Control of Data Processingcited by 16
  10. 2022DPD-fVAE: Synthetic Data Generation Using Federated Variational Autoencoders With Differentially-Private Decodercited by 14
  11. 2022DP2-VAE: Differentially Private Pre-trained Variational Autoencoderscited by 13

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Protecting Sensitive Data in Image Synthesis via PAC-Private Adaptation for Diffusion Models
  2. 2023On the Inherent Privacy Properties of Discrete Denoising Diffusion Models
  3. 2024Differentially Private Fine-Tuning of Diffusion Models
  4. 2024PID: Prompt-Independent Data Protection Against Latent Diffusion Models
  5. 2023Data Forensics in Diffusion Models: A Systematic Analysis of Membership Privacy
  6. 2024Enhancing User-Centric Privacy Protection: An Interactive Framework through Diffusion Models and Machine Unlearning
  7. 2024Shake to Leak: Fine-tuning Diffusion Models Can Amplify the Generative Privacy Risk
  8. 2024Gradient-Guided Conditional Diffusion Models for Private Image Reconstruction: Analyzing Adversarial Impacts of Differential Privacy and Denoising