The key signal is the changing geometry between two image features: the pupil and the corneal glint. As the eye rotates, their relative positions shift in a systematic way, providing a measurable basis for estimating gaze direction. This relationship converts an optical measurement into information about where the eyes are directed during a visual task.
Infrared illumination supplies the optical condition needed to make the corneal reflection visible to the tracking camera. The resulting glint acts as a reference feature on the cornea, while the pupil provides the second feature used for comparison. Their paired measurement helps distinguish gaze-related geometric changes from information provided by a single image feature alone.
Calibration establishes how measured pupil-glint geometry corresponds to locations in the visual scene. By mapping the recorded relationship to gaze positions, the system can translate subsequent eye measurements into scene coordinates. This step matters because the optical geometry must be related to the particular viewing setup before gaze data can support behavioral or perceptual analysis.
In neuroscience, the method can connect gaze measurements with visual attention, oculomotor control, and perception. Eye position provides an observable behavioral signal that can be examined alongside the visual task or scene being viewed. This makes the technique useful for studying how directed looking and eye-movement control relate to perceptual behavior without requiring invasive measurement.
A typical recording sequence begins with infrared illumination of the eye, followed by camera capture of the pupil and corneal glint. The system then measures their relative geometry, applies a calibration that maps those measurements to the visual scene, and uses the resulting gaze estimates for experimental analysis. Each stage converts optical images into interpretable eye-position data.
It is useful when a study needs a noninvasive estimate of gaze direction during visual behavior. In neuroscience, researchers can use it to examine visual attention, oculomotor control, or perception. The same approach also supports behavioral research, human-computer interaction, and clinical assessment, extending its value from basic experiments to applied evaluation of how people look at visual information.