US 11,941,499 B2
Training using rendered images
Qian Lin, Palo Alto, CA (US); Augusto Cavalcante Valente, Sao Paulo (BR); Deangeli Gomes Neves, Sao Paulo (BR); and Guilherme Augusto Silva Megeto, Sao Paulo (BR)
Assigned to Hewlett-Packard Development Company, L.P., Spring, TX (US)
Appl. No. 17/762,102
Filed by Hewlett-Packard Development Company, L.P., Spring, TX (US)
PCT Filed Oct. 16, 2019, PCT No. PCT/US2019/056536
§ 371(c)(1), (2) Date Mar. 21, 2022,
PCT Pub. No. WO2021/076125, PCT Pub. Date Apr. 22, 2021.
Prior Publication US 2022/0351427 A1, Nov. 3, 2022
Int. Cl. G06N 20/00 (2019.01); G06T 7/20 (2017.01); G06T 7/73 (2017.01); G06T 11/00 (2006.01); G06V 20/64 (2022.01)
CPC G06N 20/00 (2019.01) [G06T 7/20 (2013.01); G06T 7/75 (2017.01); G06T 11/00 (2013.01); G06V 20/647 (2022.01); G06V 20/653 (2022.01); G06T 2200/04 (2013.01); G06T 2207/20081 (2013.01)] 15 Claims
OG exemplary drawing
 
1. A method, comprising:
for a set of iterations, randomly positioning a three-dimensional (3D) object model in a virtual space having randomly selected textures;
for the set of iterations, rendering a two-dimensional (2D) image of the 3D object model in the virtual space and a corresponding annotation image; and
training a machine learning model using the rendered 2D images and corresponding annotation images.