US 12,469,184 B2
System and method for unstructured lane estimation
Alexander C. Schaefer, Fremont, CA (US); Jose Felix Ladeia Rodrigues, Sunnyvale, CA (US); and Shunsho Kaku, Mountain View, CA (US)
Assigned to Woven by Toyota, Inc., Tokyo (JP)
Filed by Woven Alpha, Inc., Tokyo (JP)
Filed on May 18, 2023, as Appl. No. 18/199,008.
Prior Publication US 2024/0386619 A1, Nov. 21, 2024
Int. Cl. G06V 20/56 (2022.01); G01C 21/00 (2006.01); G01C 21/36 (2006.01); G06T 7/13 (2017.01); G06T 11/00 (2006.01); G06V 10/44 (2022.01)
CPC G06T 11/00 (2013.01) [G01C 21/3635 (2013.01); G01C 21/3841 (2020.08); G06T 7/13 (2017.01); G06V 20/588 (2022.01); B60W 2420/403 (2013.01); G06T 2207/20076 (2013.01); G06T 2207/20084 (2013.01); G06T 2207/30241 (2013.01); G06T 2207/30256 (2013.01); G06V 10/457 (2022.01)] 20 Claims
OG exemplary drawing
 
1. A system comprising:
a processor; and
a memory in communication with the processor, the memory having instructions that, when executed by the processor, cause the processor to:
during an inference mode, output, using a machine learning model that receives as inputs an overhead representation of a region and a rasterized trace generated from sensors of a vehicle traveling in the region, probabilities that neighboring pixels of pixels forming the rasterized trace is part of a lane based on the overhead representation and the rasterized trace.