公开(公告)号：US20230040006A1

公开(公告)日：2023-02-09

申请号：US17396554

申请日：2021-08-06

Applicant: Waymo LLC

Inventor： David Joseph Weiss ,  Jeffrey Ling

IPC: B60W60/00 ,  G06N3/04 ,  G06K9/00

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for planning the future trajectory of an autonomous vehicle in an environment. In one aspect, a method comprises obtaining multiple types of scene data characterizing a scene in an environment that includes an autonomous vehicle and multiple agents; receiving route data specifying an intended route for the autonomous vehicle; for each data type, processing at least the data type using a respective encoder network to generate a respective encoding of the data type; processing a sequence of the encodings using an encoder network to generate a respective alternative representation for each data type; and processing the alternative representations using a decoder network to generate a trajectory planning output that comprises respective scores for candidate trajectories that represent predicted likelihoods that the candidate trajectory is closest to resulting in the autonomous vehicle successfully navigating the intended route.

公开(公告)号：US20230041501A1

公开(公告)日：2023-02-09

申请号：US17396560

申请日：2021-08-06

Applicant: Waymo LLC

Inventor： David Joseph Weiss ,  Jeffrey Ling ,  Adam Edward Bloniarz ,  Cole Gulino

IPC: G06N3/08 ,  G06K9/00 ,  G06T7/20

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for training a policy neural network. In one aspect, a method for training a policy neural network configured to receive a scene data input and to generate a policy output to be followed by a target agent comprises: maintaining a set of training data, the set of training data comprising (i) training scene inputs and (ii) respective target policy outputs; at each training iteration: generating additional training scene inputs; generating a respective target policy output for each additional training scene input using a trained expert policy neural network that has been trained to receive an expert scene data input comprising (i) data characterizing the current scene and (ii) data characterizing a future state of the target agent; updating the set of training data; and training the policy neural network on the updated set of training data.

公开(公告)号：US12030523B2

公开(公告)日：2024-07-09

申请号：US17396554

申请日：2021-08-06

Applicant: Waymo LLC

Inventor： David Joseph Weiss ,  Jeffrey Ling

IPC: B60W60/00 ,  G06N3/045 ,  G06V20/58

CPC classification number: B60W60/0011 ,  B60W60/00274 ,  G06N3/045 ,  G06V20/584 ,  B60W2554/20 ,  B60W2554/402

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for planning the future trajectory of an autonomous vehicle in an environment. In one aspect, a method comprises obtaining multiple types of scene data characterizing a scene in an environment that includes an autonomous vehicle and multiple agents; receiving route data specifying an intended route for the autonomous vehicle; for each data type, processing at least the data type using a respective encoder network to generate a respective encoding of the data type; processing a sequence of the encodings using an encoder network to generate a respective alternative representation for each data type; and processing the alternative representations using a decoder network to generate a trajectory planning output that comprises respective scores for candidate trajectories that represent predicted likelihoods that the candidate trajectory is closest to resulting in the autonomous vehicle successfully navigating the intended route.