Vergence Matching
Also known as Threshold-based Vergence Matching, Trigger-assisted Vergence Matching
Candidate objects near a head ray move slowly back and forth in depth, and the system selects the one whose motion matches the user's vergence eye movements.
How it works
A cone cast from the head (or hand/gaze) pre-selects the 2 or 4 nearest candidates within 10°. Each candidate oscillates in depth with a unique phase (harmonic motion, 2° vergence amplitude, 1 s cycle) and is scaled to keep a constant visual angle. Pearson correlation between the object's depth motion and the relative pupil positions (no gaze calibration needed) identifies the attended object. Threshold-based variant selects hands-free when correlation stays above 0.8; trigger-assisted variant highlights the best match and the user confirms with a controller button. In a study (N=12, HTC Vive Pro Eye) targets as small as 0.25° were selectable at 0.2–30 m, with completion times of about 3.2–3.9 s; trigger-assisted with 2 candidates was fastest, most accurate and preferred.
Papers
- Ludwig Sidenmark, Christopher Clarke, Joshua Newn, Mathias N. Lystbæk, Ken Pfeuffer, Hans Gellersen (2023). Vergence Matching: Inferring Attention to Objects in 3D Environments for Gaze-Assisted Selection. Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems (CHI '23) doi:10.1145/3544548.3580685