US 11,854,209 B2
Artificial intelligence using convolutional neural network with hough transform
Alexander Vladimirovich Sheshkus, Stary Oskol (RU); Dmitry Petrovich Nikolaev, Moscow (RU); Vladimir L'vovich Arlazarov, Moscow (RU); and Vladimir Viktorovich Arlazarov, Moscow (RU)
Assigned to Smart Engines Service, LLC, Moscow (RU)
Filed by Smart Engines Service, LLC, Moscow (RU)
Filed on Mar. 20, 2023, as Appl. No. 18/123,737.
Application 18/123,737 is a continuation of application No. 17/237,539, filed on Apr. 22, 2021, granted, now 11,636,608.
Claims priority of application No. RU2020134599 (RU), filed on Oct. 21, 2020.
Prior Publication US 2023/0245320 A1, Aug. 3, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G06T 7/168 (2017.01); G06N 3/09 (2023.01); G06N 3/048 (2023.01); G06N 3/0464 (2023.01)
CPC G06T 7/168 (2017.01) [G06N 3/048 (2023.01); G06N 3/0464 (2023.01); G06N 3/09 (2023.01); G06T 2207/20021 (2013.01); G06T 2207/20061 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01); G06T 2207/30256 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising using at least one hardware processor to:
store a neural network comprising three or more layers, a Hough Transform (HT) layer, and a Transposed Hough Transform (THT) layer, arranged such that a first subset of at least one of the three or more layers precede the HT layer, a second subset of at least one of the three or more layers follow the HT layer and precede the THT layer, and a third subset of at least one of the three or more layers follow the THT layer, wherein the HT layer converts an output of the first subset from a first space into a second space, and wherein the THT layer converts an output of the second subset from the second space into the first space; and,
for each of a plurality of input images, apply the neural network to the input image to produce an output image representing a result of an image-processing task in computer vision.