/Patent Tracker
GooglePatent applicationAudio & voiceInterest 8/10

Google patents head-worn device that cleans up voice using bone vibration

Google describes a method for a wearable device that combines a standard microphone with a sensor touching the user's head. A machine learning model uses the head's physical vibrations and a digital profile of the user's voice to rebuild their speech, effectively ignoring background noise.

Why it mattersThis could make voice commands and calls much clearer in loud environments like busy streets or construction sites.

In the filing’s own words

A method includes receiving, from a microphone, an audio signal that represents vibrations of an ambient medium induced by (i) an utterance of a user and (ii) ambient noise. The method also includes receiving, from a vibration sensor in contact with a head of the user, an inertial signal that represents vibrations of the head induced by the utterance. The method additionally includes obtaining a voice embedding that represents a voice of the user. The method further includes generating, by a machine learning model and based on (i) the audio signal, (ii) the inertial signal, and (iii) the voice embedding, a synthesized waveform that represents the utterance in the voice of the user and independently of the ambient noise. The method yet further includes outputting the synthesized waveform.

What exactly they claim (claim 1)

1. A computer-implemented method comprising: receiving, from a microphone, an audio signal that represents vibrations of an ambient medium induced by (i) an utterance of a user and (ii) ambient noise; receiving, from a vibration sensor in contact with a head of the user, an inertial signal that represents vibrations of the head induced by the utterance; obtaining a voice embedding that represents a voice of the user; generating, by a machine learning (ML) model and based on (i) the audio signal, (ii) the inertial signal, and (iii) the voice embedding, a synthesized waveform that represents the utterance in the voice of the user and independently of the ambient noise; and outputting the synthesized waveform.

Facts

Company
Google (Google LLC)
Official title
Machine Learning Based Robust Voice Communication Via Head-Worn Device
Publication number
US20260279329A1
Application number
19480258
Filed
May 9, 2023
Published
Sep 17, 2026
Inventors
4
Claims
20

Classified under

G10L 13/02 Physics & computingG02B 27/017 Physics & computingG06N 20/00 Physics & computingG10L 21/0208 Physics & computingG02B 2027/0178 Physics & computing
Read the full filing (official PDF) ↗ Open in the Patent Tracker

More from Google in Audio & voice

Summaries are simplified from the filing’s own abstract and first claim. A patent filing is not a product announcement.
Patent data and drawings: U.S. Patent and Trademark Office (USPTO). Nebula is not affiliated with or endorsed by the USPTO.