Privacy-Preserving Eye-tracking Using Deep Learning

TL;DR

This paper investigates the privacy risks in deep learning models trained on face images for eye-tracking, demonstrating that such models have a low likelihood of leaking identifiable facial features, thus preserving data privacy.

Contribution

It introduces a method to assess privacy leakage in deep models trained on face data, showing that the model maintains data privacy with reasonable confidence.

Findings

Low likelihood of facial feature memorization in the model

Deep network preserves training data privacy effectively

Method applicable to different models and conditions

Abstract

The expanding usage of complex machine learning methods like deep learning has led to an explosion in human activity recognition, particularly applied to health. In particular, as part of a larger body sensor network system, face and full-body analysis is becoming increasingly common for evaluating health status. However, complex models which handle private and sometimes protected data, raise concerns about the potential leak of identifiable data. In this work, we focus on the case of a deep network model trained on images of individual faces. Full-face video recordings taken from 493 individuals undergoing an eye-tracking based evaluation of neurological function were used. Outputs, gradients, intermediate layer outputs, loss, and labels were used as inputs for a deep network with an added support vector machine emission layer to recognize membership in the training data. The inference attack method and associated mathematical analysis indicate that there is a low likelihood of unintended memorization of facial features in the deep learning model. In this study, it is showed that the named model preserves the integrity of training data with reasonable confidence. The same process can be implemented in similar conditions for different models.