mlsanityData to 14 Oct 2026

Bounding the Invertibility of Privacy-preserving Instance Encoding using Fisher Information

arXiv 2305.04146 ↗First posted 6 May 2023cs.LG cs.CRcited by 13

Not on the map. Why?

Abstract (arXiv)

Privacy-preserving instance encoding aims to encode raw data as feature vectors without revealing their privacy-sensitive information. When designed properly, these encodings can be used for downstream ML applications such as training and inference with limited privacy risk. However, the vast majority of existing instance encoding schemes are based on heuristics and their privacy-preserving properties are only validated empirically against a limited set of attacks. In this paper, we propose a theoretically-principled measure for the privacy of instance encoding based on Fisher information. We show that our privacy measure is intuitive, easily applicable, and can be used to bound the invertibility of encodings both theoretically and empirically.

Key facts. Cited by 13 papers in the map (data through 14 Oct 2026). Builds on 39 of its 82 references that are in the map. First posted on arXiv 6 May 2023.

Builds on

39 of 82 references.

References · 39

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  3. 2018GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understandingcited by 4,899
  4. 2019DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lightercited by 4,307
  5. 2016Gaussian Error Linear Units (GELUs)cited by 3,502
  6. 2016Deep Learning with Differential Privacycited by 2,728
  7. 2019Certified Adversarial Robustness via Randomized Smoothingcited by 1,340
  8. 2017Neural Collaborative Filteringcited by 1,111
  9. 2017Deep Image Priorcited by 1,098
  10. 2017Renyi Differential Privacycited by 650
  11. 2018Federated Learning for Mobile Keyboard Predictioncited by 572
  12. 2018Exploiting Unintended Feature Leakage in Collaborative Learningcited by 520
27 more
  1. 2018Certified Robustness to Adversarial Examples with Differential Privacycited by 468
  2. 2018Lipschitz regularity of deep neural networks: analysis and efficient estimationcited by 415
  3. 2014Understanding Deep Image Representations by Inverting Themcited by 403
  4. 2018Split learning for health: Distributed deep learning without sharing raw patient datacited by 395
  5. 2019Sliced Score Matching: A Scalable Approach to Density and Score Estimationcited by 384
  6. 2020SplitFed: When Federated Learning Meets Split Learningcited by 310
  7. 2015Inverting Visual Representations with Convolutional Networkscited by 170
  8. 2020Unleashing the Tiger: Inference Attacks on Split Learningcited by 105
  9. 2020InstaHide: Instance-hiding Schemes for Private Distributed Learningcited by 90
  10. 2019Split Learning for collaborative deep learning in healthcarecited by 73
  11. 2020NoPeek: Information leakage reduction to share activations in distributed deep learningcited by 66
  12. 2017Gradient Estimators for Implicit Modelscited by 58
  13. 2022Bounding Training Data Reconstruction in Private (Deep) Learningcited by 46
  14. 2022ResSFL: A Resistance Transfer Framework for Defending Model Inversion Attack in Split Federated Learningcited by 36
  15. 2021Measuring Data Leakage in Machine-Learning Models with Fisher Informationcited by 33
  16. 2021Auto-Split: A General Framework of Collaborative Edge-Cloud AIcited by 30
  17. 2021Practical Defences Against Model Inversion Attacks for Split Neural Networkscited by 29
  18. 2021DP-Image: Differential Privacy for Image Data in Feature Spacecited by 23
  19. 2020Is Private Learning Possible with Instance Encoding?cited by 22
  20. 2022PAC Privacy: Automatic Privacy Measurement and Control of Data Processingcited by 19
  21. 2022Analyzing Privacy Leakage in Machine Learning via Multiple Hypothesis Testing: A Lesson From Fanocited by 17
  22. 2022SplitNets: Designing Neural Architectures for Efficient Distributed Computing on Head-Mounted Systemscited by 14
  23. 2021NeuraCrypt: Hiding Private Health Data via Random Neural Networks for Public Trainingcited by 11
  24. 2019Interpretable Complex-Valued Neural Networks for Privacy Protectioncited by 10
  25. 2021NeuraCrypt is not privatecited by 10
  26. 2017Estimation Efficiency Under Privacy Constraintscited by 8
  27. 2019Obfuscation via Information Density Estimationcited by 5

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 2; 2024: 1; 2025: 5; 2026: 5.

13 papers.

Sort:

By area:

  1. 2023On the Trustworthiness Landscape of State-of-the-art Generative Models: A Survey and Outlookcited by 15
  2. 2025Theoretical Insights in Model Inversion Robustness and Conditional Entropy Maximization for Collaborative Inference Systemscited by 6
  3. 2023Approximating ReLU on a Reduced Ring for Efficient MPC-based Private Inferencecited by 2
  4. 2024The Fundamental Limits of Least-Privilege Learningcited by 2
  5. 2025A Taxonomy of Attacks and Defenses in Split Learningcited by 2
  6. 2025Differentially Private 2D Human Pose Estimationcited by 1
  7. 2025CrypTorch: PyTorch-based Auto-tuning Compiler for Machine Learning with Multi-party Computationcited by 1
  8. 2026Uncertainty-Aware Gaussian Map for Vision-Language Navigationcited by 1
  9. 2025InfoDecom: Decomposing Information for Defending Against Privacy Leakage in Split Inferencecited by 0
  10. 2026Lightweight User-Personalization Method for Closed Split Computingcited by 0
  11. 2026Spectrum-Aware Bounds on Invertibility for Privacy-Enhancing Instance Encodingcited by 0
  12. 2026What Gradients Add to Text Leakage in Split Language Models, Counted per Token and per Documentcited by 0
1 more
  1. 2026From Pixels to Privacy-Sensitive Features: An Information-Theoretic Framework for Reconstruction Boundscited by 0