US 11,682,193 B2
System and method for real-time supervised machine learning in on-site environment
Chiung-Lin Chen, Beijing (CN); and Meng Yao, Beijing (CN)
Assigned to Lingdong Technology (Beijing) Co. Ltd., Beijing (CN)
Filed by Lindong Technology (Beijing) Co. Ltd., Beijing (CN)
Filed on Jul. 19, 2021, as Appl. No. 17/379,122.
Application 17/379,122 is a continuation of application No. 16/305,007, granted, now 11,092,968, previously published as PCT/CN2018/115556, filed on Nov. 15, 2018.
Prior Publication US 2021/0389775 A1, Dec. 16, 2021
This patent is subject to a terminal disclaimer.
Int. Cl. G06K 9/62 (2022.01); G06V 10/774 (2022.01); G05D 1/00 (2006.01); G05D 1/02 (2020.01); G06N 3/08 (2023.01); G06V 10/82 (2022.01); G06V 20/10 (2022.01)
CPC G06V 10/774 (2022.01) [G05D 1/0088 (2013.01); G05D 1/0212 (2013.01); G05D 1/0246 (2013.01); G06N 3/08 (2013.01); G06V 10/82 (2022.01); G06V 20/10 (2022.01); G05D 2201/0216 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01)] 20 Claims
OG exemplary drawing
 
1. An automatic guided vehicle (AGV) comprising:
a mobile base configured to operate the AGV in a self-navigation mode;
a camera configured to capture image data of at least one object; and
a main control module configured to:
receive the image data from the camera;
execute a recognition neural network program operable to recognize the at least one object in the image data;
execute a supervisor program under user control, the supervisor program configured to receive the image data and to recognize at least one marker attached to the at least one object in the image data, the supervisor program further configured to produce a supervised outcome in which the at least one object to which the at least one marker is attached is a target and is associated with a category; and
use the supervised outcome to adjust at least one weight of at least one node in the recognition neural program.