Privatization of Synthetic Gaze: Attenuating State Signatures in Diffusion-Generated Eye Movements

2026-01-28

AI Summary

This work proposes a diffusion model–based approach to synthesize eye-tracking data that preserves the functional characteristics of real eye movements while explicitly attenuating their association with subjective internal states such as fatigue and cognitive load, thereby mitigating privacy risks. By modeling eye-movement trajectories and analyzing their correlation with self-reported states, the study demonstrates for the first time that diffusion-generated data can maintain high fidelity to real data while significantly reducing statistical dependence on sensitive internal states. The method effectively balances data utility and privacy preservation, offering a viable solution for sharing and deploying eye-tracking data in privacy-sensitive applications.

Technology Category

• Machine Learning: Privacy
• Computer Vision: Diffusion Models for Vision
• Natural Language Processing: Ethics — Bias, Fairness, Transparency & Privacy

Application Category

• Security and Privacy: Data transparency and provenance
• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Responsible Web: Data and user privacy-enhancing technologies for the Web

Abstract

The recent success of deep learning (DL) has enabled the generation of high-quality synthetic gaze data. However, such data also raises privacy concerns because gaze sequences can encode subjects'internal states, like fatigue, emotional load, or stress. Ideally, synthetic gaze should preserve the signal quality of real recordings and remove or attenuate state-related, privacy-sensitive attributes. Many recent DL-based generative models focus on replicating real gaze trajectories and do not explicitly consider subjective reports or the privatization of internal states. However, in this work, we consider a recent diffusion-based gaze synthesis approach and examine correlations between synthetic gaze features and subjective reports (e.g., fatigue and related self-reported states). Our result shows that these correlations are trivial, which suggests the generative approach suppresses state-related features. Moreover, synthetic gaze preserves necessary signal characteristics similar to those of real data, which supports its use for privacy-preserving gaze-based applications.

Problem

• synthetic gaze
• privacy
• internal states
• eye movements
• privatization

Innovation

• diffusion models
• synthetic gaze
• privacy-preserving
• internal state attenuation
• eye movement generation