Decouple-and-Sample: Protecting sensitive information in task agnostic data release
Not on the map. Why?
Abstract (arXiv)
We propose sanitizer, a framework for secure and task-agnostic data release. While releasing datasets continues to make a big impact in various applications of computer vision, its impact is mostly realized when data sharing is not inhibited by privacy concerns. We alleviate these concerns by sanitizing datasets in a two-stage process. First, we introduce a global decoupling stage for decomposing raw data into sensitive and non-sensitive latent representations. Secondly, we design a local sampling stage to synthetically generate sensitive information with differential privacy and merge it with non-sensitive latent features to create a useful representation while preserving the privacy. This newly formed latent information is a task-agnostic representation of the original dataset with anonymized sensitive information. While most algorithms sanitize data in a task-dependent manner, a few task-agnostic sanitization techniques sanitize data by censoring sensitive information. In this work, we show that a better privacy-utility trade-off is achieved if sensitive information can be synthesized privately. We validate the effectiveness of the sanitizer by outperforming state-of-the-art baselines on the existing benchmark tasks and demonstrating tasks that are not possible using existing techniques.
Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Builds on 43 of its 76 references that are in the map. First posted on arXiv 17 Mar 2022.
Builds on
43 of 76 references.
References · 43
- 2015Deep Residual Learning for Image Recognitioncited by 51,809
- 2014Adam: A Method for Stochastic Optimizationcited by 36,960
- 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
- 2013Auto-Encoding Variational Bayescited by 3,980
- 2016Deep Learning with Differential Privacycited by 2,682
- 2015Variational Inference with Normalizing Flowscited by 2,161
- 2019Generating Diverse High-Fidelity Images with VQ-VAE-2cited by 1,376
- 2018Mitigating Unwanted Biases with Adversarial Learningcited by 663
- 2020NVAE: A Deep Hierarchical Variational Autoencodercited by 635
- 2018Disentangling by Factorisingcited by 614
- 2019Federated Learning with Differential Privacy: Algorithms and Performance Analysiscited by 528
- 2017Age Progression/Regression by Conditional Adversarial Autoencodercited by 426
31 more
- 2020Fairness in Machine Learning: A Surveycited by 369
- 2018Differentially Private Generative Adversarial Networkcited by 235
- 2018Isolating Sources of Disentanglement in Variational Autoencoderscited by 167
- 2021Reading Race: AI Recognises Patient's Racial Identity In Medical Imagescited by 155
- 2019Classification Accuracy Score for Conditional Generative Modelscited by 134
- 2020DP-CGAN: Differentially Private Synthetic Data and Label Generationcited by 118
- 2021Benchmarking Differentially Private Synthetic Data Generation Algorithmscited by 80
- 2020Fairness by Learning Orthogonal Disentangled Representationscited by 76
- 2018Towards Privacy-Preserving Visual Recognition via Adversarial Training: A Pilot Studycited by 48
- 2019Mitigating Information Leakage in Image Representations: A Maximum Entropy Approachcited by 48
- 2016Learning Residual Images for Face Attribute Manipulationcited by 40
- 2017A Hybrid Deep Learning Architecture for Privacy-Preserving Mobile Analyticscited by 30
- 2018Deep Private-Feature Extractioncited by 30
- 2020PrivacyNet: Semi-Adversarial Networks for Multi-attribute Face Privacycited by 30
- 2017Context-Aware Generative Adversarial Privacycited by 27
- 2020DISCO: Dynamic and Invariant Sensitive Channel Obfuscation for deep neural networkscited by 23
- 2019Context-Aware Local Differential Privacycited by 18
- 2020TIPRDC: Task-Independent Privacy-Respecting Data Crowdsourcing Framework for Deep Learning with Anonymized Intermediate Representationscited by 17
- 2016Minimax Filter: Learning to Preserve Privacy from Inference Attackscited by 16
- 2018Utility-Optimized Local Differential Privacy Mechanisms for Distribution Estimationcited by 11
- 2020InfoScrub: Towards Attribute Privacy by Targeted Obfuscationcited by 10
- 2018Anonymizing k-Facial Attributes via Adversarial Perturbationscited by 8
- 2018Generative Adversarial Privacycited by 8
- 2020Image Obfuscation for Privacy-Preserving Machine Learningcited by 7
- 2019Representation Learning with Statistical Independence to Mitigate Biascited by 6
- 2019FlowSAN: Privacy-enhancing Semi-Adversarial Networks to Confound Arbitrary Face-based Gender Classifierscited by 5
- 2019Shredder: Learning Noise Distributions to Protect Inference Privacycited by 5
- 2021Unsupervised Information Obfuscation for Split Inference of Neural Networkscited by 5
- 2021Task-aware Privacy Preservation for Multi-dimensional Datacited by 5
- 2017DEEProtect: Enabling Inference-based Access Control on Mobile Sensing Applicationscited by 2
- 2020Adversarial representation learning for synthetic replacement of private attributescited by 2
Built on by
Citing papers per year: 2023: 3; 2024: 0; 2025: 1.
4 papers.