Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Deep Learning-Based Visual Fatigue Detection Using Eye Gaze Patterns in VR

Numan Zafar, Johnathan Locke, Shafique Ahmad Chaudhry

TL;DR

This work addresses visual fatigue in VR by leveraging continuous eye-gaze trajectories collected from the GazeBaseVR dataset. It evaluates six deep time-series classifiers on five VR tasks, segmenting gaze data into short windows to detect fatigue nonintrusively via cyclopean gaze features. EKYT and related models achieve up to 94% accuracy in fatigue detection, with results varying by task and window length; gaze variance and subjective ratings provide corroborating evidence of fatigue-related changes in eye behavior. The findings support adaptive, fatigue-aware VR systems and underscore the potential and limitations of self-contained gaze-based fatigue monitoring for immersive applications.

Abstract

Prolonged exposure to virtual reality (VR) systems leads to visual fatigue, impairs user comfort, performance, and safety, particularly in high-stakes or long-duration applications. Existing fatigue detection approaches rely on subjective questionnaires or intrusive physiological signals, such as EEG, heart rate, or eye-blink count, which limit their scalability and real-time applicability. This paper introduces a deep learning-based study for detecting visual fatigue using continuous eye-gaze trajectories recorded in VR. We use the GazeBaseVR dataset comprising binocular eye-tracking data from 407 participants across five immersive tasks, extract cyclopean eye-gaze angles, and evaluate six deep classifiers. Our results demonstrate that EKYT achieves up to 94% accuracy, particularly in tasks demanding high visual attention, such as video viewing and text reading. We further analyze gaze variance and subjective fatigue measures, indicating significant behavioral differences between fatigued and non-fatigued conditions. These findings establish eye-gaze dynamics as a reliable and nonintrusive modality for continuous fatigue detection in immersive VR, offering practical implications for adaptive human-computer interactions.

Deep Learning-Based Visual Fatigue Detection Using Eye Gaze Patterns in VR

TL;DR

This work addresses visual fatigue in VR by leveraging continuous eye-gaze trajectories collected from the GazeBaseVR dataset. It evaluates six deep time-series classifiers on five VR tasks, segmenting gaze data into short windows to detect fatigue nonintrusively via cyclopean gaze features. EKYT and related models achieve up to 94% accuracy in fatigue detection, with results varying by task and window length; gaze variance and subjective ratings provide corroborating evidence of fatigue-related changes in eye behavior. The findings support adaptive, fatigue-aware VR systems and underscore the potential and limitations of self-contained gaze-based fatigue monitoring for immersive applications.

Abstract

Prolonged exposure to virtual reality (VR) systems leads to visual fatigue, impairs user comfort, performance, and safety, particularly in high-stakes or long-duration applications. Existing fatigue detection approaches rely on subjective questionnaires or intrusive physiological signals, such as EEG, heart rate, or eye-blink count, which limit their scalability and real-time applicability. This paper introduces a deep learning-based study for detecting visual fatigue using continuous eye-gaze trajectories recorded in VR. We use the GazeBaseVR dataset comprising binocular eye-tracking data from 407 participants across five immersive tasks, extract cyclopean eye-gaze angles, and evaluate six deep classifiers. Our results demonstrate that EKYT achieves up to 94% accuracy, particularly in tasks demanding high visual attention, such as video viewing and text reading. We further analyze gaze variance and subjective fatigue measures, indicating significant behavioral differences between fatigued and non-fatigued conditions. These findings establish eye-gaze dynamics as a reliable and nonintrusive modality for continuous fatigue detection in immersive VR, offering practical implications for adaptive human-computer interactions.
Paper Structure (22 sections, 3 equations, 3 figures, 8 tables)

This paper contains 22 sections, 3 equations, 3 figures, 8 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Dataset
  4. Task 1: Vergence (VRG)
  5. Task 2: Smooth pursuit task (PUR)
  6. Task 3: Video viewing (VID)
  7. Task 4: Reading text (TEX)
  8. Task 5: Random saccade task (RAN)
  9. Experiment
  10. EKYT
  11. FCN
  12. InceptionTime
  13. MCDCNN
  14. TCN
  15. TLE-Net
  16. ...and 7 more sections

Figures (3)

  • Figure 1: ROC curves for the Eye-Gaze based visual detection using Deep Classifiers.
  • Figure 2: Eye-gaze positional variance over time across different task types under conditions of fatigue and non-fatigue.
  • Figure 3: Variability in eye-gaze orientation over time across different task types under conditions of fatigue and non-fatigue.