Remote gaze tracking system for 3D environments.

Area of Science:

• Human-Computer Interaction
• Computer Vision
• Biomedical Engineering

Background:

• Traditional eye tracking systems are categorized into remote (less intrusive, for 2D screens) and mobile (more intrusive, for 3D environments).
• Remote systems typically require users to stabilize their head, limiting their application in naturalistic 3D settings.
• Existing mobile systems, like eye tracking glasses, are more intrusive and can affect user behavior.

Purpose of the Study:

• To adapt remote gaze tracking technology, originally designed for 2D screens, for use in three-dimensional (3D) environments.
• To develop a non-intrusive eye tracking solution that does not require head stabilization.
• To enable accurate gaze point estimation in 3D scenes using a remote setup.

Main Methods:

• Developed a remote gaze tracking system that maps 3D gaze points onto 2D images from a scene camera.
• The system compensates for natural, small head movements, eliminating the need for chin rests or bite bars.
• The eye tracker is located remotely from the subject, maintaining a non-intrusive observation.

Main Results:

• The system successfully tracks gaze points in a 3D environment using a remote setup.
• It demonstrates accuracy in estimating gaze points within one degree in the scene camera.
• Performance is maintained over a wide range of user head positions, compensating for movement.

Conclusions:

• Remote gaze tracking systems can be effectively extended to 3D environments.
• The developed system offers a non-intrusive and accurate method for 3D eye tracking.
• This technology has potential applications in research and interaction within complex, real-world settings.