	US 11,989,933 B2
	Polynomial convolutional neural network with late fan-out
Felix Juefei Xu, Pittsburgh, PA (US); and Marios Savvides, Pittsburgh, PA (US)
Assigned to Carnegie Mellon University, Pittsburgh, PA (US)
Appl. No. 16/976,412
Filed by CARNEGIE MELLON UNIVERSITY, Pittsburgh, PA (US)
PCT Filed Apr. 29, 2019, PCT No. PCT/US2019/029635 § 371(c)(1), (2) Date Aug. 27, 2020, PCT Pub. No. WO2019/210300, PCT Pub. Date Oct. 31, 2019.
Claims priority of provisional application 62/762,292, filed on Apr. 27, 2018.
Prior Publication US 2021/0042559 A1, Feb. 11, 2021
Int. Cl. G06V 10/82 (2022.01); G06F 18/21 (2023.01); G06F 18/2135 (2023.01); G06F 18/2413 (2023.01); G06F 18/25 (2023.01); G06N 3/045 (2023.01); G06N 3/048 (2023.01); G06N 3/08 (2023.01); G06N 3/084 (2023.01); G06V 10/44 (2022.01)

CPC G06V 10/82 (2022.01) [G06F 18/21 (2023.01); G06F 18/21355 (2023.01); G06F 18/2414 (2023.01); G06F 18/253 (2023.01); G06N 3/045 (2023.01); G06N 3/048 (2023.01); G06N 3/08 (2013.01); G06N 3/084 (2013.01); G06V 10/454 (2022.01)]

7 Claims

1. A method for training a neural network comprising, for each convolution layer in the neural network:

receiving a gradient function;

adjusting a convolutional filter based on the gradient function;

receiving an input;

convolving the convolutional filter with the input to generate a seed response map;

generating a plurality of augmented response maps based on the seed response map, wherein each element of each augmented response map is generated by applying a polynomial to one or more elements of the seed response map; and

generating a plurality of feature maps by applying an activation function to the seed response map and to each of the plurality of augmented response maps.