These smart glasses promise to turn your silent mouth movements into text
Crazy person, or efficient employee? As always, it depends on what they're wearing
Subvocal, a startup backed by Y Combinator, aims to create smart glasses capable of translating silent mouth movements into text. The company, founded by Georgia Tech computer science students Aarush Yadav and Arnav Kumar, has developed working prototypes that have been showcased in YouTube videos, including a demonstration with YC partner Dave Lieb.
The current prototype, though not yet fashionable, is significantly more streamlined than its 3D-printed predecessor. The glasses utilize low-powered radio waves to detect muscle movements and changes in facial shape as the wearer speaks or merely mouths words, which are then translated into written text. The founders claim that their device can decode open-ended speech at over 200 words per minute with a relatively low error rate even at a small dataset scale.
The core challenge the company has tackled is proving the existence of the signal and developing a method to decode it. Previous attempts using EMG electrodes on the user's face were unsuccessful, leading the team to pivot to radio frequency (RF) sensing. The final product is expected to be sleeker and more fashionable when released, potentially in early 2027 at a price of $499 for early birds.
The startup's goal is to revolutionize the way people interact with computers by allowing silent speech as an alternative to voice and keyboard input. Pre-orders are already available for $199, with the remaining balance of $300 due upon shipment.
Written by urgent.news from The Register Science's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
This story
This is one outlet's version. Read the fullest account.
