mlsanityData to 14 Oct 2026

Defending Against Gradient Inversion Attacks for Biomedical Images via Learnable Data Perturbation

arXiv 2503.16542 ↗First posted 19 Mar 2025cs.CV cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

The increasing need for sharing healthcare data and collaborating on clinical research has raised privacy concerns. Health information leakage due to malicious attacks can lead to serious problems such as misdiagnoses and patient identification issues. Privacy-preserving machine learning (PPML) and privacy-enhancing technologies, particularly federated learning (FL), have emerged in recent years as innovative solutions to balance privacy protection with data utility; however, they also suffer from inherent privacy vulnerabilities. Gradient inversion attacks constitute major threats to data sharing in federated learning. Researchers have proposed many defenses against gradient inversion attacks. However, current defense methods for healthcare data lack generalizability, i.e., existing solutions may not be applicable to data from a broader range of populations. In addition, most existing defense methods are tested using non-healthcare data, which raises concerns about their applicability to real-world healthcare systems. In this study, we present a defense against gradient inversion attacks in federated learning. We achieve this using latent data perturbation and minimax optimization, utilizing both general and medical image datasets. Our method is compared to two baselines, and the results show that our approach can outperform the baselines with a reduction of 12.5% in the attacker's accuracy in classifying reconstructed images. The proposed method also yields an increase of over 12.4% in Mean Squared Error (MSE) between the original and reconstructed images at the same level of model utility of around 90% client classification accuracy. The results suggest the potential of a generalizable defense for healthcare data.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 38 of its 76 references that are in the map. First posted on arXiv 19 Mar 2025.

Builds on

38 of 76 references.

References · 38

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  3. 2016Deep Learning with Differential Privacycited by 2,728
  4. 2019Deep Leakage from Gradientscited by 1,021
  5. 2018A Convergence Theory for Deep Learning via Over-Parameterizationcited by 787
  6. 2020Inverting Gradients -- How easy is it to break privacy in federated learning?cited by 712
  7. 2021MedMNIST v2 -- A large-scale lightweight benchmark for 2D and 3D biomedical image classificationcited by 690
  8. 2020The Future of Digital Health with Federated Learningcited by 633
  9. 2018Exploiting Unintended Feature Leakage in Collaborative Learningcited by 520
  10. 2019A Survey on Federated Learning Systems: Vision, Hype and Reality for Data Privacy and Protectioncited by 374
  11. 2021See through Gradients: Image Batch Recovery via GradInversioncited by 326
  12. 2019The Secret Revealer: Generative Model-Inversion Attacks Against Deep Neural Networkscited by 246
26 more
  1. 2021Evaluating Gradient Inversion Attacks and Defenses in Federated Learningcited by 196
  2. 2020Threats to Federated Learning: A Surveycited by 178
  3. 2019Understanding Adversarial Attacks on Deep Learning Based Medical Image Analysis Systemscited by 153
  4. 2020Provable Defense against Privacy Leakage in Federated Learning from Representation Perspectivecited by 94
  5. 2014Convolutional Neural Networks at Constrained Time Costcited by 81
  6. 2022Auditing Privacy Defenses in Federated Learning via Generative Gradient Leakagecited by 72
  7. 2020Knowledge-Enriched Distributional Model Inversion Attackscited by 67
  8. 2020The Real-World-Weight Cross-Entropy Loss Function: Modeling the Costs of Mislabelingcited by 64
  9. 2020A Framework for Evaluating Gradient Leakage Attacks in Federated Learningcited by 64
  10. 2018AttriGuard: A Practical Defense Against Attribute Inference Attacks via Adversarial Machine Learningcited by 60
  11. 2020Improving Robustness to Model Inversion Attacks via Mutual Information Regularizationcited by 52
  12. 2019Mitigating Information Leakage in Image Representations: A Maximum Entropy Approachcited by 48
  13. 2023Re-thinking Model Inversion Attacks Against Deep Neural Networkscited by 39
  14. 2022Federated Learning for Privacy Preservation in Smart Healthcare Systems: A Comprehensive Surveycited by 31
  15. 2022Bilateral Dependency Optimization: Defending Against Model-inversion Attackscited by 30
  16. 2022Learning to Invert: Simple Adaptive Attacks for Gradient Inversion in Federated Learningcited by 30
  17. 2024Privacy Leakage on DNNs: A Survey of Model Inversion Attacks and Defensescited by 30
  18. 2019Eavesdrop the Composition Proportion of Training Labels in Federated Learningcited by 27
  19. 2020Robust Aggregation for Adaptive Privacy Preserving Federated Learning in Healthcarecited by 18
  20. 2016Minimax Filter: Learning to Preserve Privacy from Inference Attackscited by 16
  21. 2024Threats and Defenses in Federated Learning Life Cycle: A Comprehensive Survey and Challengescited by 15
  22. 2020Adversarial Attacks to Machine Learning-Based Smart Healthcare Systemscited by 13
  23. 2022Federated Learning Attacks and Defenses: A Surveycited by 12
  24. 2019Shredder: Learning Noise Distributions to Protect Inference Privacycited by 5
  25. 2023RecUP-FL: Reconciling Utility and Privacy in Federated Learning via User-configurable Privacy Defensecited by 2
  26. 2024A Learning-Based Attack Framework to Break SOTA Poisoning Defenses in Federated Learningcited by 2

Built on by

Not cited yet.