US 11,836,968 B1
Systems and methods for configuring and using a multi-stage object classification and condition pipeline
Robert Winston Blanchard, San Diego, CA (US); and Neela Niranjani Vengateshwaran, Canton, MI (US)
Assigned to SAS Institute, Inc., Cary, NC (US)
Filed by SAS INSTITUTE INC., Cary, NC (US)
Filed on Aug. 24, 2023, as Appl. No. 18/237,866.
Claims priority of provisional application 63/431,277, filed on Dec. 8, 2022.
Int. Cl. G06V 10/764 (2022.01); G06T 3/40 (2006.01); G06T 7/70 (2017.01); G06T 7/11 (2017.01)
CPC G06V 10/764 (2022.01) [G06T 3/40 (2013.01); G06T 7/11 (2017.01); G06T 7/70 (2017.01); G06V 2201/07 (2022.01)] 30 Claims
OG exemplary drawing
 
1. A computer-program product embodied in a non-transitory machine-readable storage medium storing computer instructions that, when executed by one or more processors, perform operations comprising:
detecting, via a localization machine learning model, a target object within a scene based on downsampled image data of the scene;
identifying, via the one or more processors, a likely position of the target object within original image data of the scene based on object position data computed by the localization machine learning model;
extracting, from the original image data of the scene, a target sub-image containing the target object based on the likely position of the target object;
classifying, via an object classification machine learning model, the target object to a probable object class of a plurality of distinct object classes based on a target image resolution of the target sub-image, wherein the plurality of distinct object classes includes an out-of-scope object class and one or more in-scope object classes;
routing, via the one or more processors, the target image resolution of the target sub-image to a target object-condition machine learning classification model of a plurality of distinct object-condition machine learning classification models based on a mapping between the plurality of distinct object classes and the plurality of distinct object-condition machine learning classification models;
classifying, via the target object-condition machine learning classification model, the target object to a probable object-condition class of a plurality of distinct object-condition classes; and
displaying, via a graphical user interface, a representation of the target object in association with the probable object-condition class based on the classifying of the target object via the target object-condition machine learning classification model.