FocusFlow: Gaze-Depth Interaction in VR

FocusFlow

Gaze is an intuitive and efficient way to interact in VR, but current methods rarely make use of the depth control built into our visual system. This project turns visual depth into a hands-free input dimension. We developed a binocular visual-depth detection algorithm and a layer-based interface built around a “Virtual Window,” which keeps gaze-depth interaction intuitive and robust even though depth precision drops at farther distances. To help newcomers learn this unfamiliar input, we designed two learning strategies with different visual cues. A user study with 24 participants confirmed the usability of the Virtual Window and showed that adaptive visual cues help users build muscle memory for gaze-depth control.

Publications