US 11,810,343 B2
Artificial intuition based visual data extraction for distributed systems
Eran Pereg, Tel-Aviv (IL); and Reuven Sabi, Ashdod (IL)
Assigned to Asio Advanced Control Solutions Ltd, Kfar-Monash (IL)
Filed by Asio Advanced Control Solutions Ltd, Kfar-Monash (IL)
Filed on May 11, 2022, as Appl. No. 17/741,497.
Claims priority of provisional application 63/186,816, filed on May 11, 2021.
Prior Publication US 2022/0366683 A1, Nov. 17, 2022
Int. Cl. G06K 9/62 (2022.01); G06V 10/98 (2022.01); G06T 7/70 (2017.01); G06V 20/50 (2022.01); G06V 10/764 (2022.01); G06V 10/70 (2022.01); G06F 18/25 (2023.01); G06V 10/75 (2022.01); G06V 10/94 (2022.01); G08B 13/196 (2006.01); G06V 10/20 (2022.01); G06V 20/52 (2022.01); G06V 20/40 (2022.01); G06F 18/20 (2023.01)
CPC G06V 10/98 (2022.01) [G06F 18/20 (2023.01); G06F 18/251 (2023.01); G06F 18/254 (2023.01); G06T 7/70 (2017.01); G06V 10/255 (2022.01); G06V 10/70 (2022.01); G06V 10/75 (2022.01); G06V 10/764 (2022.01); G06V 10/768 (2022.01); G06V 10/95 (2022.01); G06V 20/44 (2022.01); G06V 20/50 (2022.01); G06V 20/52 (2022.01); G08B 13/19604 (2013.01); G08B 13/19663 (2013.01)] 19 Claims
OG exemplary drawing
 
1. A system for visually identifying anomaly events using, at an edge node, a context based Machine Learning model (ML) that increases a performance of the edge node, comprising:
said edge node comprising at least one processor configured to execute a program code, the program code comprising:
code instructions to apply, by the at least one processor of said edge node, a limited resources classifier to a plurality of images captured by at least one imaging sensor deployed to monitor a certain scene relating to a certain area to classify at least one object detected in at least one image of the plurality of images;
code instructions to apply, by the at least one processor of said edge node, a trained context based Machine Learning (ML) model to classification data generated by the limited resources classifier to compute, by said trained context based ML model applied at said edge node, an anomaly score for at least one potential anomaly event relating to the at least one detected object based on at least one contextual attribute associated with the certain scene, wherein said at least one contextual attribute is derived by said trained context based ML model applied at said edge node, wherein said classification data comprises metadata relating to one or more objects, actions, events, conditions, situations, states, and scenarios detected in the plurality of images; and
code instructions to transmit the at least one image from the edge node to a remote server in cases where the anomaly score exceeds a first threshold, wherein the remote server is configured to apply at least one high performance visual analysis tool to visually analyze the at least one image in order to identify the at least one potential anomaly event.