US 12,423,583 B2
Selecting annotations for training images using a neural network
Xiaosong Wang, Rockville, MD (US); Ziyue Xu, Reston, VA (US); Dong Yang, Pocatello, ID (US); Lickkong Tam, Santa Clara, CA (US); and Daguang Xu, Potomac, MD (US)
Assigned to NVIDIA Corporation, Santa Clara, CA (US)
Filed by NVIDIA Corporation, Santa Clara, CA (US)
Filed on Jun. 1, 2020, as Appl. No. 16/889,670.
Prior Publication US 2021/0374547 A1, Dec. 2, 2021
Int. Cl. G06N 3/084 (2023.01); G06F 18/214 (2023.01); G06F 18/24 (2023.01); G06F 40/20 (2020.01); G06N 3/047 (2023.01)
CPC G06N 3/084 (2013.01) [G06F 18/214 (2023.01); G06F 18/24 (2023.01); G06F 40/20 (2020.01); G06N 3/047 (2023.01)] 34 Claims
OG exemplary drawing
 
1. One or more processors comprising circuitry to:
use one or more neural networks to:
obtain data as input to train the one or more neural networks;
generate a set of weights for the one or more neural networks by processing a portion of the data;
compute, using the set of weights, a set of features representing characteristics corresponding to a label set of the portion;
compute weights for one or more labels from the label set using the set of features; and
select at least one label from the one or more labels based on the weights to update the one or more neural networks; and
cause the one or more labels to be compared with one or more objects within one or more images identified by the one or more neural networks.