Granted Patent B2

US 12,705,109 · App. 18/203,535 · Granted Aug 11, 2026

Inventors: David D. Lee (Palo Alto, CA); Andrew Augustine Wajs (Haarlem, NL)

Assignee: Scenera, Inc.

G06F9/5072G06F9/542G06F18/251G06F18/285G06N20/00G06V10/454G06V10/764G06V10/82G06V10/94G06V20/52H04N23/80H04N23/90G06V40/161G06V2201/10

Abstract

Claims (45)

1 . A method for enabling application-configured awareness of spaces, the method comprising:

receiving, via an application programming interface (API), requests from applications for different monitorings of spaces; and

configuring a plurality of non-human technological entities to implement workflows for the requested monitorings of spaces, the entities including cameras that view the monitored spaces, wherein the workflows include:

the cameras capturing images of the monitored spaces;

artificial intelligence and/or machine learning (AI/ML) entities detecting events from the captured images;

generating SceneMarks with attributes that are descriptive of the detected events;

transmitting the SceneMarks between the AI/ML entities;

at least one AI/ML entity performing analysis of SceneMarks received via the transmission between the AI/ML entities; and adding information to the attributes of at least one of the received SceneMarks based on said analysis, and/or detecting an event based on said analysis and generating a new SceneMark for said detected event; and

the workflows contextualize the captured images from the SceneMarks to provide awareness of situations in the monitored spaces.

2 . The method of claim 1 wherein the at least one AI/ML entity performs analysis of received SceneMarks to detect an anomaly in the situation in the monitored space.

3 . The method of claim 2 wherein the at least one AI/ML entity detects the anomaly based on a sequence of received SceneMarks.

4 . The method of claim 2 wherein the anomaly is one of: an unexpected occupancy of the space, an unusual movement of a person through the space, an unusual interaction between people in the space, an unexpected object in the space, or an unexpected condition for the space.

5 . The method of claim 2 wherein the at least one AI/ML entity detects the anomaly based on comparing received SceneMarks with SceneMarks produced by a normal situation in the monitored space.

6 . The method of claim 2 wherein the at least one AI/ML entity detects the anomaly based on comparing received SceneMarks with SceneMarks predicted for a normal situation in the monitored space.

7 . The method of claim 1 wherein the at least one AI/ML entity determines that a person or object identified in two different SceneMarks are the same person or object.

8 . The method of claim 7 wherein the at least one AI/ML entity determines that the person or object in the two different SceneMarks are the same person or object, based on attributes of the two different SceneMarks.

9 . The method of claim 7 wherein the at least one AI/ML entity determines that the person or object in the two different SceneMarks are the same person or object, based on timestamps of the two different SceneMarks and a known proximity of cameras capturing the images that generated the two different SceneMarks.

10 . The method of claim 1 wherein the workflow further includes:

automatically triggering an action based on a sequence of SceneMarks that are indicative of a predefined situation in the monitored space.

11 . The method of claim 1 wherein the workflow further includes:

classifying received SceneMarks into different predefined categories; and

automatically triggering different actions based on the category.

12 . The method of claim 1 wherein the workflow further includes:

generating a text description of the situation in the monitored space, based on received SceneMarks.

13 . The method of claim 12 wherein a generative pre-trained transformer entity generates the text description.

14 . The method of claim 12 wherein the workflow further includes:

returning the text description of the situation to the requesting application.

15 . The method of claim 12 wherein generating the text description comprises:

generating labels based on the received SceneMarks; and

matching the generated labels against predefined labels that describe different situations.

16 . The method of claim 1 wherein the workflow further includes:

accessing SceneMarks stored in a SceneMark database, wherein the workflows contextualize the captured images from SceneMarks including the stored SceneMarks.

17 . The method of claim 16 wherein a generative pre-trained transformer entity formulates a query to access the SceneMarks stored in the SceneMark database.

18 . The method of claim 16 wherein a generative pre-trained transformer entity formulates a natural language response based on SceneMarks returned from the SceneMark database in response to a query.

19 . The method of claim 1 wherein the SceneMarks include links to the captured images, and the SceneMarks are transmitted between entities but the captured images are not transmitted between entities.

20 . A system comprising:

a plurality of applications that make requests for different monitorings of spaces;

a plurality of non-human technological entities, the entities including cameras that view the monitored spaces and further including artificial intelligence and/or machine learning (AI/ML) entities;

a service that receives the requests and configures the entities to implement workflows for the requested monitorings of spaces, wherein the workflows include:

the cameras capturing images of the monitored spaces;

the AI/ML entities detecting events from the captured images;

generating SceneMarks with attributes that are descriptive of the detected events;

transmitting the SceneMarks between the AI/ML entities;

at least one AI/ML entity performing analysis of SceneMarks received via the transmission between the AI/ML entities; and adding information to the attributes of at least one of the received SceneMarks based on said analysis, and/or detecting an event based on said analysis and generating a new SceneMark for said detected event; and

the workflows contextualize the captured images from the SceneMarks to provide awareness of situations in the monitored spaces.

Assignments (1)

Continuity (6)

References Cited (88)