Samsung Patents a Two-Step Gesture System for AR Wearables That Skips the Button-Tap

By Patentlyze Team · Updated Jun 5, 2026

Tapping a tiny floating button in AR is harder than it sounds — Samsung's new patent sidesteps the problem entirely by watching where you point before you commit to any action.

What Samsung's proximity-first AR input actually does

Imagine you're wearing AR glasses and a handful of virtual buttons are floating in your field of view. The annoying part: they're small, your hand isn't perfectly steady, and you have to nail the exact button to trigger the right thing. Samsung's patent describes a system that takes the pressure off that final tap.

Here's the flow: you point or gesture near a general area — not necessarily at a specific button — and the device logs that as a first signal. If you then make a second gesture in the same spot within a short window of time, the wearable figures out which nearby virtual object you were probably aiming for and shows you a visual cue for what to do next.

That third step — the visual cue — tells you how to confirm the action using a different sensor entirely (think an eye-blink, a finger pinch, or a voice command). It's a forgiving, multi-stage input model that doesn't require you to be pixel-perfect in mid-air.

How the three-input pipeline resolves your gesture intent

The patent describes a three-phase input pipeline for a wearable device (think AR glasses or a mixed-reality headset) that relies on at least two different sensors working in sequence.

Phase 1 — Proximity detection: The first sensor picks up a user input at a location that isn't directly on any virtual object. This is the "rough aim" — you're near the buttons but not touching any of them. The system registers this as a meaningful, intentional signal rather than noise.

Phase 2 — Intent confirmation: If a second input arrives at the same location within a preset duration (a configurable time window), the device interprets the two inputs together. It then identifies the virtual object or objects that are spatially closest to that location — essentially inferring which UI element you were probably pointing at.

Phase 3 — Guided confirmation: The wearable displays a visual object — an on-screen overlay or indicator — that shows the user what input a second sensor is waiting for to complete the action. That second sensor could be a camera watching for an eye gesture, a microphone, a biosignal sensor, or anything else the device supports.

The separation of "locate" and "confirm" across two sensors is the core idea: the first sensor handles spatial targeting, the second handles intentional execution.

What this means for hands-free AR control on Galaxy devices

Mid-air interaction with AR UIs is notoriously imprecise — your arm gets tired, your aim drifts, and false positives are annoying. By splitting the input into a rough-aim phase and a confirm phase that uses a different sensor, Samsung's approach could make AR interfaces feel much less like a frustrating game of darts. You don't have to be precise on the first move — the system meets you halfway.

This fits neatly into Samsung's broader wearables push, particularly as it expands into the AR/XR space alongside its Galaxy Ring and rumored glasses projects. A more forgiving input model could also reduce the need for precise hand-tracking hardware, potentially lowering the cost bar for capable AR wearables.

Editorial commentary on a publicly published patent application. Not legal advice.