US 11,887,046 B1
Updating machine learning systems using previous output data
Christopher Andrew Stephens, Issaquah, WA (US); Alexander Clark Prater, Seattle, WA (US); Alexander Michael McNamara, Seattle, WA (US); Sridhar Boyapati, Sammamish, WA (US); David Echevarria Ignacio, Seattle, WA (US); David William Bettis, Seattle, WA (US); Korwin Jon Smith, Seattle, WA (US); Kevin Alexander Lee, Seattle, WA (US); Aaron Craig Thompson, Seattle, WA (US); Gary Paolo Raden, Seattle, WA (US); Sudarshan Narasimha Raghavan, Snoqualmie, WA (US); Dilip Kumar, Seattle, WA (US); and Félix Joseph Étienne Pageau, Seattle, WA (US)
Assigned to AMAZON TECHNOLOGIES, INC., Seattle, WA (US)
Filed by AMAZON TECHNOLOGIES, INC., Seattle, WA (US)
Filed on Jul. 31, 2020, as Appl. No. 16/945,289.
Application 16/945,289 is a continuation of application No. 14/502,280, filed on Sep. 30, 2014, granted, now 10,733,565.
This patent is subject to a terminal disclaimer.
Int. Cl. G06Q 10/087 (2023.01); G06Q 10/08 (2023.01); G06T 7/00 (2017.01); G06V 40/10 (2022.01); H04N 7/18 (2006.01)
CPC G06Q 10/087 (2013.01) [G06Q 10/08 (2013.01); G06T 7/0004 (2013.01); G06V 40/10 (2022.01); H04N 7/181 (2013.01); G06T 2207/30108 (2013.01); G06T 2207/30196 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
accessing image data associated with a facility;
determining occurrence of an event at the facility involving a first user handling an item at the facility;
determining, based on the image data, one or more possible item identifiers of the item handled by the first user;
determining a confidence level associated with each of the one or more possible item identifiers;
determining each of the confidence levels is less than a minimum threshold;
based on the determining that the each of the confidence levels is less than the minimum threshold, generating inquiry data associated with a selection of one of the one or more possible item identifiers, wherein the inquiry data comprises:
at least a portion of the image data, and
the one or more possible item identifiers;
sending the inquiry data to a device associated with a second user;
causing a user interface to be presented to the second user, the user interface including the inquiry data;
accessing response data, responsive to the inquiry data, the response data associated with the selection of the one of the one or more possible item identifiers; and
designating the image data and at least a portion of the response data as training data for a machine learning system.