mlsanityData to 7 Oct 2026

Rethinking Privacy Preserving Deep Learning: How to Evaluate and Thwart Privacy Attacks

arXiv 2006.11601 ↗First posted 20 Jun 2020cs.LG cs.CR cs.DCcited by 32

Not on the map. Why?

Abstract (arXiv)

This paper investigates capabilities of Privacy-Preserving Deep Learning (PPDL) mechanisms against various forms of privacy attacks. First, we propose to quantitatively measure the trade-off between model accuracy and privacy losses incurred by reconstruction, tracing and membership attacks. Second, we formulate reconstruction attacks as solving a noisy system of linear equations, and prove that attacks are guaranteed to be defeated if condition (2) is unfulfilled. Third, based on theoretical analysis, a novel Secret Polarization Network (SPN) is proposed to thwart privacy attacks, which pose serious challenges to existing PPDL methods. Extensive experiments showed that model accuracies are improved on average by 5-20% compared with baseline mechanisms, in regimes where data privacy are satisfactorily protected.

Key facts. Cited by 32 papers in the map (data through 7 Oct 2026). First posted on arXiv 20 Jun 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 6; 2022: 6; 2023: 11; 2024: 7; 2025: 1; 2026: 1.

32 papers.

Sort:

  1. 2021Robbing the Fed: Directly Obtaining Private Data in Federated Learning with Modified Modelscited by 106
  2. 2021Gradient Inversion with Generative Image Priorcited by 101
  3. 2022Survey on Federated Learning Threats: concepts, taxonomy on attacks and defences, experimental study and challengescited by 81
  4. 2021CAFE: Catastrophic Data Leakage in Vertical Federated Learningcited by 76
  5. 2023GIFD: A Generative Gradient Inversion Method with Feature Domain Optimizationcited by 45
  6. 2022A Survey on Gradient Inversion: Attacks, Defenses and Future Directionscited by 39
  7. 2023LOKI: Large-scale Data Reconstruction Attack against Federated Learning through Model Manipulationcited by 35
  8. 2022Reconstructing Training Data from Model Gradient, Provablycited by 27
  9. 2023Privacy and Fairness in Federated Learning: on the Perspective of Trade-offcited by 26
  10. 2023A Survey of What to Share in Federated Learning: Perspectives on Model Utility, Privacy Leakage, and Communication Efficiencycited by 21
  11. 2022A Hybrid Self-Supervised Learning Framework for Vertical Federated Learningcited by 17
  12. 2021Understanding Training-Data Leakage from Gradients in Neural Networks for Image Classificationcited by 16
20 more
  1. 2021Federated Deep Learning with Bayesian Privacycited by 13
  2. 2024The Federation Strikes Back: A Survey of Federated Learning Privacy Attacks, Defenses, Applications, and Policy Landscapecited by 13
  3. 2022Concealing Sensitive Samples against Gradient Leakage in Federated Learningcited by 11
  4. 2023Surrogate Model Extension (SME): A Fast and Accurate Weight Update Attack on Federated Learningcited by 10
  5. 2023FedPass: Privacy-Preserving Vertical Federated Deep Learning with Adaptive Obfuscationcited by 9
  6. 2024Revisiting Gradient Pruning: A Dual Realization for Defending against Gradient Attackscited by 7
  7. 2023Theoretically Principled Federated Learning for Balancing Privacy and Utilitycited by 6
  8. 2023Understanding Deep Gradient Leakage via Inversion Influence Functionscited by 6
  9. 2024Towards Eliminating Hard Label Constraints in Gradient Inversion Attackscited by 4
  10. 2021Quantifying and Localizing Usable Information Leakage from Neural Network Gradientscited by 2
  11. 2023Evaluating Privacy Leakage in Split Learningcited by 2
  12. 2023Privacy Preserving Federated Learning with Convolutional Variational Bottleneckscited by 2
  13. 2024Leak and Learn: An Attacker's Cookbook to Train Using Leaked Data from Federated Learningcited by 2
  14. 2026Enhancing Gradient Inversion Attacks in Federated Learning via Hierarchical Feature Optimizationcited by 2
  15. 2022Analysing Training-Data Leakage from Gradients through Linear Systems and Gradient Matchingcited by 1
  16. 2023FedDMF: Privacy-Preserving User Attribute Prediction using Deep Matrix Factorizationcited by 0
  17. 2024Adaptive Hybrid Masking Strategy for Privacy-Preserving Face Recognition Against Model Inversion Attackcited by 0
  18. 2024FedAdOb: Privacy-Preserving Federated Deep Learning with Adaptive Obfuscationcited by 0
  19. 2024Privacy Attack in Federated Learning is Not Easy: An Experimental Studycited by 0
  20. 2025Byzantine Outside, Curious Inside: Reconstructing Data Through Malicious Updatescited by 0