Do Commercial Video See-Through AR HMDs with Foveated Rendering and Autofocus Cameras Improve 3D Pointing? Misc

										
@misc{ranapurwala2026vacautofocus,
title = {Do Commercial Video See-Through AR HMDs with Foveated Rendering and Autofocus Cameras Improve 3D Pointing?},
author = {Ranapurwala, Amaan and Wu, Junyuan and Mutasim, Aunnoy K and Batmaz, Anil Ufuk and Zielasko, Daniel and Stuerzlinger, Wolfgang and Barrera Machuca, Mayra D.},
booktitle = {Symposium on Spatial User Interaction},
publisher = {ACM},
series = {SUI '26},
howpublished = {Poster},
year = {2026},
month = {Oct},
numpages = {2},
doi = {https://doi.org/10.1145/3822518.3841191},
pdf = {papers/vacautofocus.pdf},
abstract = {Commercial Video See-Through Augmented Reality (VST AR) head-mounted displays (HMDs) increasingly combine eye tracking, foveated rendering, and camera-based focus mechanisms. We evaluate two commercial VST AR HMDs with/without autofocus VST cameras in a within-subjects study to identify their effect on 3D interaction. Participants performed an ISO 9241-411 pointing task under three depth conditions: no vergence-accommodation conflict (No VAC), constant VAC, and varying VAC. Our results show that both HMDs had a weak Fitts' law fit in the No VAC and Varying VAC conditions, suggesting that angular task difficulty alone does not fully explain pointing performance in gaze-contingent VST AR systems.},
keywords = {video see-through, autofocus, VAC},
note = {To appear}
}