Reconstruction of Privacy-Sensitive Data from Protected Templates

05/08/2019

∙

In this paper, we address the problem of data reconstruction from privacy-protected templates, based on recent concept of sparse ternary coding with ambiguization (STCA). The STCA is a generalization of randomization techniques which includes random projections, lossy quantization, and addition of ambiguization noise to satisfy the privacy-utility trade-off requirements. The theoretical privacy-preserving properties of STCA have been validated on synthetic data. However, the applicability of STCA to real data and potential threats linked to reconstruction based on recent deep reconstruction algorithms are still open problems. Our results demonstrate that STCA still achieves the claimed theoretical performance when facing deep reconstruction attacks for the synthetic i.i.d. data, while for real images special measures are required to guarantee proper protection of the templates.

READ FULL TEXT

Reconstruction of Privacy-Sensitive Data from Protected Templates

A Linear Reconstruction Approach for Attribute Inference Attacks against Synthetic Data

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

Privacy-Preserving SVM Computing by Using Random Unitary Transformation

TAPAS: a Toolbox for Adversarial Privacy Auditing of Synthetic Data

Privacy-Preserving Synthetic Educational Data Generation

Privacy-Preserving Near Neighbor Search via Sparse Coding with Ambiguation

Covariance's Loss is Privacy's Gain: Computationally Efficient, Private and Accurate Synthetic Data

Reconstruction of Privacy-Sensitive Data from Protected Templates

Related Research

A Linear Reconstruction Approach for Attribute Inference Attacks against Synthetic Data

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

Privacy-Preserving SVM Computing by Using Random Unitary Transformation

TAPAS: a Toolbox for Adversarial Privacy Auditing of Synthetic Data

Privacy-Preserving Synthetic Educational Data Generation

Privacy-Preserving Near Neighbor Search via Sparse Coding with Ambiguation

Covariance's Loss is Privacy's Gain: Computationally Efficient, Private and Accurate Synthetic Data