CPC G06F 18/214 (2023.01) [G06F 3/14 (2013.01); G06F 18/40 (2023.01); G06N 20/00 (2019.01); G06V 20/00 (2022.01)] | 20 Claims |
1. A method, comprising:
training a machine learning model, wherein training the machine learning model comprises:
receiving a plurality of images of a training object corresponding to a target object;
receiving, via a user interface, an annotation associated with the training object;
determining a plurality of characteristics of the training object; and
associating, based at least in part on the annotation, the plurality of characteristics with an object class corresponding to the training object;
receiving, from at least one imaging device positioned in a waste processing facility, imaging data associated with a portion of a volume of solid waste;
receiving, from a separator control assembly that controls an operation of a separator, telemetry data indicating one or more of a weight of an object being held by the separator or a resistance force applied against the separator;
identifying, using the trained machine learning model and based at least in part on the imaging data and the telemetry data, the target object, wherein the target object is disposed within the volume of solid waste; and
providing, to the separator control assembly, an output associated with the target object,
wherein the output is to facilitate removal, by the separator, of the target object from the volume of solid waste, and
wherein the output includes a display instruction to cause a display device of the separator control assembly to display a representation of the target object within, and distinguished from, the volume of solid waste.
|