US 11,989,642 B2
Future object trajectory predictions for autonomous machine applications
Ruben Villegas, Ann Arbor, MI (US); Alejandro Troccoli, San Jose, CA (US); Iuri Frosio, San Jose, CA (US); Stephen Tyree, University City, MO (US); Wonmin Byeon, Santa Clara, CA (US); and Jan Kautz, Lexington, MA (US)
Assigned to NVIDIA Corporation, Santa Clara, CA (US)
Filed by NVIDIA Corporation, Santa Clara, CA (US)
Filed on Sep. 26, 2022, as Appl. No. 17/952,866.
Application 17/952,866 is a continuation of application No. 16/564,978, filed on Sep. 9, 2019, granted, now 11,514,293.
Claims priority of provisional application 62/729,659, filed on Sep. 11, 2018.
Prior Publication US 2023/0088912 A1, Mar. 23, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G06N 3/044 (2023.01); B60W 40/02 (2006.01); G06N 3/045 (2023.01); G06N 3/08 (2023.01)
CPC G06N 3/044 (2023.01) [B60W 40/02 (2013.01); G06N 3/08 (2013.01); G06N 3/045 (2023.01)] 20 Claims
OG exemplary drawing
 
1. A method comprising:
generating, based at least on sensor data, one or more state representations by populating one or more first spaces of the one or more state representations to indicate one or more first locations of a first object within an environment and one or more second spaces of the one or more state representations to indicate one or more second locations of one or more second objects within the environment;
generating, based at least on the one or more state representations, one or more spatial representations by populating one or more third spaces of the one or more spatial representations using the one or more first locations and one or more fourth spaces of the one or more spatial representations using the one or more second locations, the one or more spatial representations indicative of the one or more first locations of the first object relative to the one or more second locations of one or more second objects within the environment; and
determining, based at least on the one or more state representations and the one or more spatial representations, at least one of a maneuver associated with the first object or a future location associated with the first object.