Computationally-efficient human-identifying smart assistant computer

Patent metadata

Field	Value
Publication number	US-2018232902-A1
Application number	US-201715636422-A
Country	US
Kind code	A1
Filing date	Jun 28, 2017
Priority date	Feb 14, 2017
Publication date	Aug 16, 2018
Grant date	—

Abstract

A computationally-efficient method for a smart assistant computer to track a human includes receiving data from one or more sensors configured to monitor a physical environment. The data is computer-analyzed to recognize presence of a human in the physical environment, and upon confirming an identity of the human, a first level of computational resources of the smart assistant computer is dedicated to track the human. Upon failing to confirm the identity of the human while a known user is present, a second level of computational resources of the smart assistant computer, greater than the first level, is dedicated to determine the identity of the human. Upon failing to confirm the identity of the human while the known user is absent, a third level of computational resources of the smart assistant computer, is dedicated to determine the identity of the human.

First claim

1 . A method for a smart assistant computer to track a human, the method comprising:
receiving data from one or more sensors configured to monitor a physical environment; computer-analyzing the data to recognize presence of a human in the physical environment; and
upon confirming an identity of the human, dedicating a first level of computational resources of the smart assistant computer to track the human; or
upon failing to confirm the identity of the human while a known user is present, dedicating a second level of computational resources of the smart assistant computer, greater than the first level of computational resources, to determine the identity of the human; or
upon failing to confirm the identity of the human while the known user is absent, dedicating a third level of computational resources of the smart assistant computer, greater than the second level of computational resources, to determine the identity of the human.

2 . The method of claim 1 , where dedicating the first level of computational resources to track the human includes operating at least one of the one or more sensors at a reduced sampling frequency.

3 . The method of claim 1 , where dedicating the second level of computational resources to track the human includes operating at least one of the one or more sensors at an increased sampling frequency.

4 . The method of claim 1 , where dedicating the second level of computational resources to determine the identity of the human includes applying a first set of image processing algorithms to one or both of a visible-light image of the human captured by a visible-light camera of the one or more sensors or an infrared image of the human captured by an infrared camera of the one or more sensors.

5 . The method of claim 4 , where the first set of image processing algorithms includes one or more contrast adjustment algorithms.

6 . The method of claim 5 , where a contrast adjustment algorithm of the one or more contrast adjustment algorithms includes identifying lowest and highest pixel luminance values L i and H i in an infrared image of the physical environment, such values ranging from 0 to 255, and changing L i and H i to contrast-adjusted values L CA and H CA , where L CA is given by

MIN


(

220
,

Hi
+

(

255
-
Hi

)

2

)

.

and H CA is given by

MAX


(

30
,

Li
2

)

,

7 . The method of claim 4 , where dedicating the third level of computational resources to determine the identity of the human includes applying a second set of image processing algorithms to the visible-light image of the human, the second set of image processing algorithms requiring more processor operations than the first set of image processing algorithms.

8 . The method of claim 1 , where dedicating the third level of computational resources to determine the identity of the human includes prioritizing processor threads associated with human identification over processor threads associated with other tasks.

9 . The method of claim 1 , where dedicating the third level of computational resources to determine the identity of the human includes dedicating one or more user interface resources of the smart assistant computer to satisfy a security check.

10 . The method of claim 1 , where dedicating the third level of computational resources to determine the identity of the human includes sending an identification query to an electronic device associated with the known user.

11 . The method of claim 1 , further comprising sending a location query to an electronic device associated with the known user.

12 . The method of claim 1 , where one to all of the first, second, and third levels of computational resources of the smart assistant computer are dynamically scaled based on one or more security factors.

13 . The method of claim 12 , where a current level of computational resources dedicated to identifying the human is scaled up when the smart assistant computer is in a high security mode.

14 . The method of claim 12 , where a current level of computational resources dedicated to identifying the human is scaled up during a predetermined period.

15 . The method of claim 12 , where a current level of computational resources dedicated to identifying the human is scaled up when an unaccompanied protected user is present.

16 . The method of claim 1 , where computer-analyzing the data to recognize presence of the human includes detecting motion in the physical environment.

17 . The method of claim 16 , where motion detection begins after receiving an indication that an entrance door in the physical environment has opened.

18 . The method of claim 1 , where computer-analyzing the data to recognize presence of the human includes detecting presence of a known portable computing device in the physical environment.

19 . A smart assistant computer, comprising:
a logic processor; and a storage device holding instructions executable by the logic processor to:
receive data from one or more sensors configured to monitor a physical environment;
computer-analyze the data to recognize presence of a human in the physical environment; and
upon confirming an identity of the human, dedicate a first level of computational resources of the smart assistant computer to track the human; or
upon failing to confirm the identity of the human while a known user is present, dedicate a second level of computational resources of the smart assistant computer, greater than the first level of computational resources, to determine the identity of the human; or
upon failing to confirm the identity of the human while the known user is absent, dedicate a third level of computational resources of the smart assistant computer, greater than the second level of computational resources, to determine the identity of the human.

20 . A method for a smart assistant computer to track a human, the method comprising:
receiving data from one or more sensors configured to monitor a physical environment; computer-analyzing the data to recognize presence of a human in the physical environment;
if presence of the human is recognized in the environment, computer-analyzing the data to identify the human;
if the human is identified, dedicating a first level of computational resources of the smart assistant computer to track the human;
if the human is not identified, determining if a known user is presen

Assignees

• Microsoft Technology Licensing Llc

Inventors

• Albadawi Haithem
• Liu Zongyi

Classifications

• H04N23/611

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

• G06F18/29

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

• G06F18/251

  of input or preprocessed data · 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