User registration for intelligent assistant computer

Patent metadata

Field	Value
Publication number	US-2018232201-A1
Application number	US-201715682425-A
Country	US
Kind code	A1
Filing date	Aug 21, 2017
Priority date	Feb 14, 2017
Publication date	Aug 16, 2018
Grant date	—

Abstract

Registration of a person with an intelligent assistant computer includes obtaining one or more image frames captured via one or more cameras that depict an initially unregistered person. Facial recognition data for the initially unregistered person is extracted from the one or more image frames. A spoken command to register the initially unregistered person is received via one or more microphones. Upon determining that the spoken command originated from the registered person having the pre-established registration privilege, the initially unregistered person is registered as a newly registered person by associating one or more additional privileges with the facial recognition data in a person profile for the newly registered person.

First claim

1 . A method performed by a computing system for registering a person with an intelligent assistant computer, the method comprising:
obtaining one or more image frames captured via one or more cameras that depict an initially unregistered person; extracting facial recognition data for the initially unregistered person from the one or more image frames; receiving a spoken command to register the initially unregistered person via one or more microphones; determining that the spoken command originated from a registered person having a pre-established registration privilege; and upon determining that the spoken command originated from the registered person having the pre-established registration privilege, registering the initially unregistered person as a newly registered person by associating one or more additional privileges with the facial recognition data in a person profile for the newly registered person.

2 . The method of claim 1 , wherein the one or more image frames are captured after receiving the spoken command to register the initially unregistered person.

3 . The method of claim 2 , further comprising:
responsive to receiving the spoken command to register the initially unregistered person, directing the initially unregistered person to position its face within a field of view of the one or more cameras to capture the one or more image frames for facial recognition.

4 . The method of claim 3 , wherein directing the initially unregistered person includes one or more of outputting an audible direction via an audio speaker and/or outputting a visual direction via a graphical display device.

5 . The method of claim 1 , further comprising:
responsive to receiving the spoken command to register the initially unregistered person, directing the initially unregistered person to speak one or more words or phrases; obtaining one or more audio segments captured via one or more microphones that include the one or more words or phrases spoken by the initially unregistered person; extracting speaker recognition data from the one or more audio segments for the initially unregistered person; and associating the speaker recognition data with the person profile for the newly registered person.

6 . The method of claim 1 , wherein the spoken command to register the initially unregistered person is received after the one or more image frames are captured via the one or more cameras.

7 . The method of claim 6 , further comprising:
storing the one or more image frames in a data storage system prior to receiving the spoken command to register the initially registered person; retrieving the one or more image frames from the data storage system; presenting the one or more image frames via a graphical display device for review by the registered person; and wherein the spoken command is received during or after presentation of the one or more image frames.

8 . The method of claim 7 , wherein the one or more image frames are presented after the initially unregistered person leaves a field of view of the one or more cameras.

9 . The method of claim 7 , wherein the one or more image frames are presented responsive to another command initiated by the registered person.

10 . The method of claim 1 , wherein the one or more image frames form part of one or more video segments captured via the one or more cameras; and
wherein the method further comprises:
identifying a speaking activity of the initially unregistered person within the one or more video segments;
obtaining one or more audio segments, captured via one or more microphones, that are time-matched to the one or more video segments;
extracting speaker recognition data for the initially unregistered person from the one or more audio segments based on one or more spoken words or phrases that correspond to the speaking activity of the initially unregistered person; and
associating the speaker recognition data with the person profile.

11 . The method of claim 10 , further comprising:
receiving a subsequent spoken command to perform one or more operations via one or more microphones; determining that the subsequent spoken command originated from the newly registered person having the one or more additional privileges based on the speaker recognition data; and performing an operation of the one or more operations that is permitted by the one or more additional privileges responsive to the spoken command.

12 . The method of claim 1 , wherein the spoken command forms part of a spoken phrase that originated from the registered person that further includes a person identifier for the newly registered person; and
wherein the method further comprises associating the person identifier with the person profile for the newly registered person.

13 . The method of claim 1 , wherein the spoken command forms part of a spoken phrase that originated from the registered person that further includes a privileges identifier for the newly registered person that identifies the one or more additional privileges associated with the person profile;
wherein each privilege of the one or more additional privileges permits one or more operations, not previously permitted prior to registration, to be performed by the intelligent assistant computer responsive to a command originating from the newly registered person.

14 . The method of claim 13 , wherein the privilege identifier indicates whether the newly registered person is permitted to register other initially unregistered persons.

15 . A computing system, comprising:
one or more cameras to capture image data; one or more microphones to capture audio data; one or more computing devices implementing an intelligent assistant service configured to:
obtain one or more image frames captured via the one or more cameras that depict an initially unregistered person;
extract facial recognition data for the initially unregistered person from the one or more image frames;
receive a spoken command to register the initially unregistered person via the one or more microphones;
determine that the spoken command originated from a registered person having a pre-established registration privilege; and
upon determining that the spoken command originated from the registered person having the pre-established registration privilege, register the initially unregistered person as a newly registered person by associating one or more additional privileges with the facial recognition data in a person profile for the newly registered person stored in a data storage system of the one or more computing devices.

16 . The computing system of claim 15 , wherein the one or more image frames are captured after receiving the spoken command to register the initially unregistered person; and
wherein the intelligent assistant service is further configured to:
responsive to receiving the spoken command to register the initially unregistered person, direct the initially unregistered person to position its face within a field of view of the one or more cameras to capture the one or more image frames for facial recognition.

17 . The computing system of claim 15 , wherein the intelligent assistant service is further configured to:
responsive to receiving the spoken command to register the initially unregistered person, direct the initially unregistered person to speak one or more words or phrases; obtain one or more audio segments captured via the one or more microphones that include the one or more words or phrases spoken by the initially unregistered person;
extract speaker recognition data from

Assignees

• Microsoft Technology Licensing Llc

Inventors

• Holtmann Michelle Lynn

Classifications

• H04N23/611

  where the recognised objects include parts of the human body · CPC title

• G06F18/41

  Interactive pattern learning with a human teacher · CPC title

• G06N3/044

  Recurrent networks, e.g. Hopfield networks · CPC title

• H04N23/90

  Arrangement of cameras or camera modules, e.g. multiple cameras in TV studios or sports stadiums · CPC title

• G06F18/29

  Graphical models, e.g. Bayesian networks · CPC title