/Patent Tracker
GooglePatent applicationSecurity & privacyInterest 7/10

Google patents local face recognition for smart home cameras

This filing describes a system where security cameras and doorbells identify faces directly on the device rather than sending video to the cloud. It uses local software to match faces against a stored library of known people, allowing the system to alert owners if a stranger is detected while keeping raw image data private.

Why it mattersIt offers a way for smart home users to get personalized alerts about who is at their door without sacrificing the privacy of their video footage.

In the filing’s own words

The present document describes systems and methods for on-device person-recognition and provision of intelligent alerts. The system includes a decentralized multi-camera system for on-device facial recognition. A device (e.g., security camera, video doorbell) captures images/video of a person, processes input image frames, detects face images, filters static faces, and aligns a rotation of the face to be upright and frontal. The device then filters low-quality face images and/or images having a large portion of the face occluded. The device computes a face embedding, compares it against a set of locally stored reference embeddings, and sends matching results to cloud services, which, based on the matching result, notifies the device owner whether the observed person is a known person or a stranger. Face detection and recognition computations are performed on-device, not at the cloud. No sensitive information is transmitted off device and privacy is thus preserved.

What exactly they claim (claim 1)

1. A method comprising: storing a library of clusters in a local memory on a first wireless network device of a decentralized multi-device system, the first wireless network device having a facial recognition system and an on-device machine-learning framework, a respective cluster including one or more exemplars that are each associated with a same face, at least one exemplar of the one or more exemplars in the respective cluster including: a reference embedding; an indication of a local version of a facial recognition model used by the on-device machine-learning framework to generate the reference embedding; and an exemplar-quality score of the at least one exemplar in relation to the facial recognition model; receiving, by the first wireless network device from an off-device service, a new reference exemplar corresponding to a face detected in an image captured by a second wireless network device of the decentralized multi-device system, the new reference exemplar including metadata information and a first quality score but does not include a face embedding; downloading, by the first wireless network device from the off-device service, a snapshot image corresponding to the new reference exemplar; generating, by the on-device machine-learning framework on the first wireless network device, a new face embedding representing the face in the downloaded snapshot image using the local version of the facial recognition model on the first wireless network device; and storing a new exemplar in the library of clusters in the local memory of the first wireless network device, the new exemplar including the new face embedding, an indication of the local version of the facial recognition model used by the first wireless network device to generate the new face embedding, and a second quality score of the new exemplar in relation to the facial recognition model.

Facts

Company
Google (Google LLC)
Official title
Systems and Methods for On-Device Person Recognition and Provision of Intelligent Alerts
Publication number
US20260268708A1
Application number
19658526
Filed
Apr 25, 2026
Published
Sep 10, 2026
Inventors
6
Claims
20

Classified under

G06V 40/172 Physics & computingG06V 10/762 Physics & computingG06V 10/95 Physics & computing
Read the full filing (official PDF) ↗ Open in the Patent Tracker

More from Google in Security & privacy

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.