US 12,444,126 B1
Neural network-based view synthesis
Jonathan Tremblay, Redmond, WA (US); Moustafa Meshry, College Park, MD (US); Charles Loop, Mercer Island, WA (US); and Towaki Takikawa, Toronto (CA)
Assigned to NVIDIA CORPORATION, Santa Clara, CA (US)
Filed by Nvidia Corporation, Santa Clara, CA (US)
Filed on Mar. 11, 2022, as Appl. No. 17/692,909.
Int. Cl. G06T 15/06 (2011.01); G06N 3/045 (2023.01); G06N 3/08 (2023.01); G06T 15/08 (2011.01); G06T 15/50 (2011.01)
CPC G06T 15/06 (2013.01) [G06N 3/045 (2023.01); G06N 3/08 (2013.01); G06T 15/08 (2013.01); G06T 15/50 (2013.01)] 31 Claims
OG exemplary drawing
 
1. One or more processors, comprising: circuitry to:
use one or more neural networks to:
encode one or more features of one or more voxels of a three-dimensional (3D) representation of an object based, at least in part, on one or more two-dimensional (2D) images of the object; and
use one or more decoders of the one or more neural networks to determine one or more optical properties of one or more voxels of the object based, at least in part, on a path of a simulated ray intersecting the one or more voxels of the object.