Context-aided identification

Patent metadata

Field	Value
Publication number	US-12322396-B2
Application number	US-202218050329-A
Country	US
Kind code	B2
Filing date	Oct 27, 2022
Priority date	Nov 3, 2021
Publication date	Jun 3, 2025
Grant date	Jun 3, 2025

Abstract

Smart devices can be configured to collect and share various forms of context data about where a user is located (e.g., location), what a user will be doing (e.g., schedule), and what a user is currently doing (e.g., activity). This context data may be combined with fingerprint data (e.g., biometrics) to help identify the fingerprint data. For example, a location of a user may help associated speech detected at that location with the user. These associations may be stored in a mapping database that can be updated over time to reduce ambiguities in identification. The mappings in the database may be used to train a machine learning model to recognize fingerprints as identities, which may be useful in applications, such as speaker identification.

First claim

The invention claimed is:

1. A method comprising:
receiving recognition information corresponding to an unknown identity;
receiving context information, the context information correlated at a time with the recognition information;
determining possible identities of the unknown identity based on the context information;
querying a database with the recognition information; and
when an association between the recognition information and one of the possible identities is identified in the database, then:
adjusting a confidence level of the association based on the context information correlated at the time with the recognition information.

2. The method according to claim 1 , wherein the recognition information includes a corresponding confidence, and the context information includes a corresponding confidence.

3. The method according to claim 2 , further comprising:
associating the recognition information with the possible identities to generate a plurality of associations between the recognition information and the possible identities; and
training a machine learning model with the plurality of associations between the recognition information and the possible identities.

4. The method according to claim 3 , wherein associating the recognition information with the possible identities includes:
mapping the recognition information and its corresponding confidence with the context information and its corresponding confidence to generate mappings; and
storing the mappings as database entries in the database.

5. The method according to claim 4 , further comprising:
updating the database based on the plurality of associations between the recognition information and the possible identities by:
determining that the association between the recognition information and one of the possible identities is not identified in the database; and
recording the association in the database with an initial confidence level.

6. The method according to claim 1 , wherein adjusting the confidence level of the association includes increasing the confidence level when the context information supports the association.

7. The method according to claim 1 , wherein adjusting the confidence level of the association includes decreasing the confidence level when the context information does not support the association.

8. The method according to claim 1 , identifying the unknown identity as the one of the possible identities when the confidence level of the association is adjusted above a threshold.

9. The method according to claim 1 , wherein the recognition information is a voice fingerprint based on audio captured by microphones of a smart eyewear.

10. The method according to claim 1 , wherein the recognition information is a face fingerprint based on images captured by a camera of a smart eyewear.

11. The method according to claim 1 , wherein the context information includes a calendar entry corresponding to the time that the recognition information was gathered.

12. The method according to claim 1 , wherein the recognition information is gathered by a smart eyewear and the context information includes locations of devices near the smart eyewear at the time that the recognition information was gathered.

13. The method according to claim 1 , wherein the database is stored locally on a smart eyewear.

14. The method according to claim 13 , further comprising triggering, by an application running on the smart eyewear, identification of the recognition information.

15. The method according to claim 14 , wherein the application is a speech-to-text transcript configured to tag a transcript by speaker based on identified voice fingerprints.

16. Augmented reality glasses, comprising:
at least one sensor configured to gather first sensor data at a first time and second sensor data at a second time, the second time after the first time;
a communication module configured to receive context information from other devices in communication with the augmented reality glasses, the context information corresponding to the first time that the first sensor data was gathered;
a memory storing a database; and
a processor communicatively coupled to the at least one sensor and the communication module that is configured by software instructions to:
receive the first sensor data;
identify the first sensor data as a fingerprint;
receive the context information;
determine possible identities of the fingerprint from the context information;
associate the fingerprint with the possible identities to generate a plurality of associations between the fingerprint and the possible identities;
updating the database with the plurality of associations between the fingerprint and the possible identities;
receive the second sensor data;
identify the second sensor data as the fingerprint;
querying the database with the fingerprint;
identify a previously recorded association of the fingerprint with one of the possible identities in the database, the previously recorded association having a confidence level;
change the confidence level of the previously recorded association in the database based on the context information; and
identify the fingerprint from a particular identity when the confidence level is above a threshold.

17. The augmented reality glasses according to claim 16 , wherein the context information includes identities corresponding to devices within a range of the augmented reality glasses.

18. The augmented reality glasses according to claim 16 , wherein the fingerprint is a voice print corresponding to speech gathered by the at least one sensor.

19. The augmented reality glasses according to claim 18 , wherein the processor is further configured to:
transcribe the speech to generate transcribed speech;
tag the transcribed speech with the particular identity in a transcript; and
display the transcript on a display of the augmented reality glasses.

20. The augmented reality glasses according to claim 16 wherein the processor is further configured to:
add an association of the fingerprint with the particular identity when the database does not include the association of the fingerprint with the particular identity.

21. The augmented reality glasses according to claim 16 , wherein:
the confidence level is increased when the context information supports the previously recorded association; or
the confidence level is decreased when the context information does not support the previously recorded association.

22. Augmented reality glasses comprising:
at least one sensor configured to capture biometric information corresponding to an unknown identity;
a communication module configured to receive context information, the context information correlated at a time with the biometric information;
a memory storing a database; and
a processor configured by software instructions to:
determine possible identities of the unknown identity based on the context information;
query the database with the biometric information;
identify a previously recorded association between the biometric information and a particular identity of the possible identities in the database, the previously recorded association having a confidence level; and
increase the confidence level of the previously recorded association in the database based on the context information correlated at the time with the biometric information.

23. The augmented reality glasses according to claim 22 , whe

Assignees

• Google Llc

Inventors

• Olwal Alex
• Du Ruofei

Classifications

• G10L17/04

  Training, enrolment or model building · CPC title

• G06V40/50

  Maintenance of biometric data or enrolment thereof · CPC title

• G06V40/70

  Multimodal biometrics, e.g. combining information from different biometric modalities · CPC title

• G06V40/172

  Classification, e.g. identification · CPC title

• H04L63/0861

  using biometrical features, e.g. fingerprint, retina-scan (cryptographic mechanisms or cryptographic arrangements for entity authentication using biological data H04L9/3231) · CPC title