Privacy-Preserving Eye Tracking

• Applicants: Professor Dr. Andreas Bulling; Professor Dr. Ralf Küsters
• Subject Area: Image and Language Processing, Computer Graphics and Visualisation, Human Computer Interaction, Ubiquitous and Wearable Computing ; Security and Dependability, Operating-, Communication- and Distributed Systems
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 548713845

Eye tracking technology has significantly advanced in recent years and, with it, has its proliferation in various applications in both research and industry. In parallel, human gaze measured through eye tracking has emerged as a rich source of information on users' activities and focus of visual attention but also on even more private information, such as users' intentions and goals, personal preferences, personality traits, and their health status. With eye tracking becoming pervasive, fundamental new challenges regarding privacy arise. However, this critical topic has so far remained under-investigated in the eye tracking research community: Privacy-preserving eye tracking has only recently started to attract attention, privacy threats are poorly understood, and effective privacy-preserving technologies as well as formal privacy solutions are largely missing. The goal of this project is to address this research gap and to advance privacy-preserving eye tracking along the whole eye tracking data processing pipeline. To this end, the project proposes a comprehensive work program at the intersection of machine learning, computer vision, and human-computer interaction on the one hand as well as privacy-preserving technologies on the other. We specifically focus on cryptographic approaches, including multi-party computation, differential privacy, and (fully) homomorphic encryption. We will use and advance these approaches for private learning-based gaze estimation and gaze data quality verification, private unlearning of gaze behaviour models, and privacy-preserving gaze-based user behaviour modelling. As such, this project aims to pioneer and lay the scientific foundations for a new generation of privacy-preserving eye tracking methods with provable privacy guarantees.

• DFG Programme: Research Grants