US 12,093,843 B2
Inferencing and learning based on sensorimotor input data
Jeffrey C. Hawkins, Atherton, CA (US); Subutai Ahmad, Palo Alto, CA (US); Yuwei Cui, Redwood City, CA (US); and Marcus Anthony Lewis, San Francisco, CA (US)
Assigned to Numenta, Inc., Redwood City, CA (US)
Filed by Numenta, Inc., Redwood City, CA (US)
Filed on Mar. 11, 2021, as Appl. No. 17/198,808.
Application 17/198,808 is a continuation of application No. 15/594,077, filed on May 12, 2017, granted, now 10,977,566.
Claims priority of provisional application 62/335,995, filed on May 13, 2016.
Prior Publication US 2021/0201181 A1, Jul. 1, 2021
This patent is subject to a terminal disclaimer.
Int. Cl. G06N 5/04 (2023.01); G06N 20/00 (2019.01)
CPC G06N 5/04 (2013.01) [G06N 20/00 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method of performing inference, comprising:
generating, by a first input processor, a first input representation indicating a first set of candidate pairs, the first set of candidate pairs indicating detection of a first feature of an unidentified object at first candidate locations associated with the first feature;
determining, by a first output processor coupled to the first input processor, a first output representation corresponding to the first input representation, the first output representation indicating a first set of candidate objects associated with the first set of pairs and including the unidentified object;
generating, by a second input processor, a second input representation indicating a second set of candidate pairs, the second set of candidate pairs indicating detection of a second feature of the unidentified object at second candidate locations associated with the second feature; and
determining, by a second output processor coupled to the second input processor, a second output representation indicating a second set of candidate objects associated with the second set of pairs and that is consistent with one or more candidate objects of the first output representation.