Overview
Developed eye-tracking software for a NASA-funded project focused on detecting fighter pilot fatigue during high-stress flight simulations. The system uses a Pupil Labs Core eye-tracking headset combined with ArUco marker detection to map gaze onto a physical cockpit display, enabling real-time fatigue analysis.
Highlights
- Built a real-time gaze mapping pipeline using OpenCV and ArUco markers to correlate eye position with specific cockpit instrument panels
- Integrated Pupil Labs Core hardware via ZMQ for low-latency gaze data streaming
- Implemented fatigue classification heuristics based on blink rate, fixation duration, and saccade velocity
- Developed a data collection pipeline compliant with university RISK/IRB protocols for human subject research
Tech Stack
PythonOpenCVArUcoPupil LabsZMQNumPyCUDA