Testing a Motion Matching Algorithm for Gaze-based HMI

Abstract

This study explores gaze-based interaction, focusing on object selection, particularly in dynamic environments. Traditional methods like dwell-time selection have limitations, prompting investigation into novel approaches such as motion matching. A pilot study was conducted to compare a motion matching algorithm with dwell-time selection, indicating a tendency towards faster selection times with motion matching. Workload metrics showed very similar results between selection mechanisms, but a small bias towards reduced user frustration and enhanced satisfaction using motion matching. Challenges remain, including the Midas touch problem and technical constraints of eye tracking technology, highlighting the need for further research to refine algorithms and address limitations. Despite challenges, motion matching represents progress towards making gaze-based interaction more accessible for widespread use.

Keywords: Gaze-based interaction, motion matching, eye tracking, object selection, HMI

DOI: 10.54941/ahfe1005648