US 12,128,895 B2
Autonomous driving device
Hyunki Seong, Busan (KR); Chanyoung Jung, Seoul (KR); Seungwook Lee, Seoul (KR); and David Hyunchul Shim, Daejeon (KR)
Assigned to SK hynix Inc., Icheon (KR); and Korea Advanced Institute of Science and Technology, Daejeon (KR)
Filed by SK hynix Inc., Icheon (KR); and Korea Advanced Institute of Science and Technology, Daejeon (KR)
Filed on Oct. 3, 2022, as Appl. No. 17/937,733.
Claims priority of application No. 10-2021-0157644 (KR), filed on Nov. 16, 2021.
Prior Publication US 2023/0150498 A1, May 18, 2023
Int. Cl. B60W 30/14 (2006.01); B60W 30/18 (2012.01); B60W 40/02 (2006.01); B60W 60/00 (2020.01); G06N 3/045 (2023.01); G06N 3/049 (2023.01)
CPC B60W 30/143 (2013.01) [B60W 30/18159 (2020.02); B60W 40/02 (2013.01); B60W 60/001 (2020.02); G06N 3/045 (2023.01); G06N 3/049 (2013.01); B60W 2556/40 (2020.02)] 11 Claims
OG exemplary drawing
 
1. An autonomous driving device, comprising:
an execution network configured to determine a target speed of a driving vehicle at a current time according to a state information history including a plurality of state information for road environment, the plurality of state information being generated at a plurality of times, respectively,
wherein the execution network includes:
a spatial attention network configured to receive the state information history and to generate feature data based on the state information history; and
a temporal attention network configured to determine the target speed of the driving vehicle by applying temporal importance to an output of the spatial attention network,
wherein each of the plurality of state information included in the state information history includes location information of surrounding vehicles, wherein the temporal attention network includes: a second attention network configured to receive an output of the spatial attention network and to provide more weight on data having more temporal importance; and an output neural network configured to generate the target speed using an output of the second attention network.