Mobile eye tracker ST ONE

ST ONE is a wearable, glasses-form-factor eye-tracking system designed for collecting high-quality eye-movement data in real-world and mobile conditions where head motion, device slippage, and interruptions are expected. Unlike many eye trackers optimized for static lab setups, ST ONE is built for naturalistic behavior, allowing rapid head movement, re-wearing between sessions, and use with corrective eyewear.

The system relies on infrared cameras and a binocular, truly 3D eye model to measure real eye movements directly, rather than estimating gaze via end-to-end deep learning. This approach improves robustness across diverse users and environments and supports consistent accuracy without frequent recalibration. A key feature is persistent calibration: once calibrated, users can remove and re-don the device without repeating calibration procedures.

ST ONE records eye-tracking data synchronized with a forward-facing scene camera, enabling contextual gaze analysis in real-world tasks. Captured signals include eye position and orientation (x, y, z), gaze point coordinates, pupil size (in millimeters), blink events, fixation events, and validity codes. Data can be streamed live or recorded for offline analysis.

For development and integration, recorded data can be exported in standard formats such as CSV and gaze video, allowing straightforward use in Python, MATLAB, or R analysis pipelines. Real-time streaming enables multimodal experiments combining eye tracking with EEG, physiological signals, motion capture, or XR systems.

Typical applications include cognitive and behavioral science, usability and UX research, XR interaction studies, training and performance analysis, and human–computer interaction experiments that require unrestricted head and body movement.

Development & Research Advice: Validate calibration stability and data quality under your expected motion conditions before running full experiments. When combining ST ONE with XR headsets or motion capture, use a shared timestamping or synchronization strategy. For real-time applications, prototype first with recorded data to tune latency and filtering before deploying live pipelines.