Related Experiment Video
Updated: Feb 26, 2026
08:06
Eye Movement Monitoring of Memory
Published on: August 15, 2010
15.2K
Web Camera Based Eye Tracking to Assess Visual Memory on a Visual Paired Comparison Task
Nicholas T Bott1, Alex Lange2, Dorene Rentz3,4
1Department of Medicine, School of Medicine, Stanford UniversityStanford, CA, United States.
Frontiers in Neuroscience
|July 14, 2017
Summary
Web cameras can accurately track eye movements for decision-making tasks, correlating well with standard eye trackers. This low-cost method offers a viable alternative for research, despite limitations in fine-grained analysis.
Area of Science:
- Ophthalmology and Vision Science
- Human-Computer Interaction
- Cognitive Neuroscience
Background:
- Web cameras are ubiquitous in smart devices, presenting opportunities for novel research applications.
- Eye movements serve as a non-invasive indicator of cognitive processes, making eye tracking a valuable research tool.
- Recording eye movements via web cameras is an emerging field with significant potential for accessible research.
Purpose of the Study:
- To introduce and evaluate a new method for conducting a visual paired comparison (VPC) task using a standard web camera.
- To assess the correlation between eye movement data captured by a high-frame-rate (60 FPS) eye tracker and a low-frame-rate (3 FPS) built-in laptop web camera.
- To determine the reliability and accuracy of web camera-based eye movement tracking for decisional tasks.
Main Methods:
- An observational study involving 54 healthy older adults across three in-clinic visits.
- Simultaneous recording of eye movements during a VPC task using a professional eye tracker and a built-in laptop web camera.
- Analysis of inter-rater reliability using Siegel and Castellan's kappa formula and correlation analysis using Pearson's r.
Main Results:
- A strong correlation (r = 0.88-0.92) was found between the 60 FPS eye tracker and the 3 FPS web camera for VPC mean novelty preference scores.
- High inter-rater agreement (κ = 0.81-0.88) was achieved for web camera scoring.
- The built-in web camera method encountered fewer data quality issues and showed strong performance across various frame rates (10, 5, and 3 FPS).
Conclusions:
- Manual scoring of a VPC task using built-in web cameras shows strong correlation with automated scoring from high-speed eye trackers.
- This web camera methodology is a cost-effective and accurate approach for tracking eye movements in decisional tasks.
- While not suitable for high-precision metrics like fixation points, built-in web cameras offer a practical solution for many eye movement research applications.
