Gaze Cone + Gaze Probe Disambiguation
Also known as Gaze+Gesture+Disambiguation (GGD), Gaze Probe
An invisible 5° cone around the gaze collects nearby or hidden objects, which then spread out in a circle so the user can look at the intended one and close the hand to select it.
How it works
The user fixates for 200 ms and swipes a hand toward the gaze point (Leap Motion). An invisible 5° gaze cone gathers every object whose centroid lies inside it. If only one object is caught it is highlighted; if several are, the rest of the scene is blurred and the candidates animate onto a circle on a single depth plane, centred on their average centroid, so none overlap. The user then looks at the intended candidate and closes the palm to confirm; swiping away cancels and opening the palm deselects. In a desktop study (N=12, Tobii EyeX + Leap Motion) against plain Gaze+Gesture and a mouse, it nearly eliminated errors under partial and full occlusion but was slower when nothing was occluded.
Papers
- Shujie Deng, Jian Chang, Jian Jun Zhang, Shi-Min Hu (2017). Gaze Modulated Disambiguation Technique for Gesture Control in 3D Virtual Objects Selection. 2017 3rd IEEE International Conference on Cybernetics (CYBCONF) doi:10.1109/CYBConf.2017.7985779