Gaze Authentication: Factors Influencing Authentication Performance

2025-09-13

AI Summary

This study systematically investigates key factors affecting gaze-based authentication performance. We employ a large-scale, self-collected dataset comprising 8,849 subjects and a 72-Hz video-based eye-tracking system to quantitatively evaluate the impact of signal quality, calibration strategies, and filtering on authentication accuracy. Our experiments—conducted on Meta Quest Pro–class hardware, using an end-to-end eye-tracking pipeline and state-of-the-art neural network architectures—reveal that: (1) fixing calibration target depth, (2) fusing calibrated and uncalibrated gaze data, and (3) enhancing raw signal quality significantly improve recognition accuracy; conversely, conventional low-pass filtering slightly degrades performance. These findings, validated under realistic conditions, identify core optimization pathways for gaze authentication and expose the limitations of standard preprocessing techniques. The work provides empirical evidence and a novel design paradigm for robust, low-calibration-dependency gaze authentication systems.

Technology Category

• Computer Vision: Motion & Tracking
• Machine Learning: Calibration & Uncertainty Quantification
• Search and Optimization: Evaluation and Analysis

Application Category

• Security and Privacy: Large-scale security measurements
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

This paper examines the key factors that influence the performance of state-of-the-art gaze-based authentication. Experiments were conducted on a large-scale, in-house dataset comprising 8,849 subjects collected with Meta Quest Pro equivalent hardware running a video oculography-driven gaze estimation pipeline at 72Hz. The state-of-the-art neural network architecture was employed to study the influence of the following factors on authentication performance: eye tracking signal quality, various aspects of eye tracking calibration, and simple filtering on estimated raw gaze. We found that using the same calibration target depth for eye tracking calibration, fusing calibrated and non-calibrated gaze, and improving eye tracking signal quality all enhance authentication performance. We also found that a simple three-sample moving average filter slightly reduces authentication performance in general. While these findings hold true for the most part, some exceptions were noted.

Problem

• Examining factors influencing gaze-based authentication performance
• Studying eye tracking calibration and signal quality effects
• Evaluating filtering impact on authentication accuracy

Innovation

• Same depth calibration targets
• Fusing calibrated and non-calibrated gaze
• Improving eye tracking signal quality