ExpertEyes: open-source, high-definition eyetracking

Area of Science:

• Biomedical Engineering
• Computer Vision
• Human-Computer Interaction

Background:

• Accurate eye tracking is crucial for research in various fields.
• Existing commercial eye trackers can be expensive and lack portability.
• Challenges exist in eye tracking within demanding recording environments.

Purpose of the Study:

• To introduce ExpertEyes, an open-source, portable, high-definition eye tracking system.
• To detail the technological innovations enabling its performance.
• To provide a cost-effective and accessible alternative to commercial systems.

Main Methods:

• Developed a novel forward eye model for simultaneous pupil and corneal reflection fitting using full ellipses.
• Utilized high-resolution eye imaging by directly recording to SD cards from uncased digital camcorders.
• Implemented offline processing for optimized parameter estimation.

Main Results:

• Achieved typical accuracy of less than 0.4° and best accuracy below 0.3°.
• Demonstrated typical precision (SD method) around 0.3° and best precision below 0.2°.
• The system's performance is comparable to or exceeds commercial eye tracking systems.

Conclusions:

• ExpertEyes provides a portable, high-definition, and accurate eye tracking solution.
• The open-source nature and detailed hardware plans promote accessibility and further development.
• The system is suitable for challenging recording environments and offers a viable alternative to expensive commercial options.