Researchers Vimal Mollyn and Chris Harrison propose GazeTags, a way to tell a wearable which object you're looking at without a camera or computer vision.
How it works: The worn device, such as smartglasses or hearables, periodically emits a narrow infrared signal aligned with the user's head direction. Only tags inside that beam wake and respond with their identity and optional payloads, such as object state or sensor readings.
Cost and power: Each tag costs roughly $2 and can run for more than a year on a coin cell battery.
Why no camera: The paper says this makes the approach very energy efficient and avoids non-trivial privacy issues.
Status: It's listed as published at UIST 2026, with the arXiv version submitted 2 Nov 2026. The abstract says the paper covers hardware design, performance and tradeoffs, and demos.
No cameras, no vision pipeline, just a beam and a tag.
Sign in to comment.
TL;DR of the paper, generated automatically.
Researchers Vimal Mollyn and Chris Harrison propose GazeTags, a way to tell a wearable which object you're looking at without a camera or computer vision.
No cameras, no vision pipeline, just a beam and a tag.