Evaluating the Vergence–Accommodation Conflict in Gaze-Based 3D Target Selection

This study investigates how the Vergence–Accommodation Conflict (VAC) influences gaze-based 3D target selection in virtual environments. While the VAC is known to degrade depth perception and manual interaction performance, its specific impact on gaze-driven input—which relies directly on oculomotor mechanisms—remained under-explored. The researchers conducted a within-subjects user study with 24 participants, comparing gaze-based pointing against controller-based raycasting. Targets were presented at six different depths, measured in diopters, to systematically vary the vergence-accommodation demand relative to the HMD's fixed focal plane.

The results demonstrate that gaze-based selection performance is not uniform across depth. As targets move further from the display's focal plane, users experience a significant decrease in selection speed and accuracy. The study confirms that the VAC negatively affects gaze-based pointing, similar to its known effects on manual interaction. Furthermore, the researchers found that a diopter-aware Fitts' law model—which accounts for the optical focus demands of the eye—provides a superior fit for the observed performance data compared to traditional linear models. This suggests that depth-dependent factors are critical for accurately predicting and designing gaze-based interactions in 3D space.

As eye tracking becomes a standard feature in modern XR headsets, gaze is increasingly used as a primary input modality. Designers often place interactive elements at varying depths to create immersive 3D interfaces. This paper provides empirical evidence that these interfaces must account for the VAC to ensure consistent interaction quality. By validating a diopter-based performance model, the study offers a practical tool for developers to predict interaction difficulty and optimize the placement of gaze-selectable targets in virtual environments.