Motion planning with implicit occupancy for autonomous systems

A system comprising:

• a computer processor;
• and
• non-transitory computer readable medium for causing the computer processor to perform operations comprising:
• obtaining, by a virtual driver of an autonomous system, a set of trajectories through a geographic region for an autonomous system,
• generating, for each trajectory in the set of trajectories, a set of points of interest in the geographic region to obtain sets of points of interest,
• quantizing the sets of points of interest to obtain a set of query points in the geographic region,
• querying an implicit decoder model with the set of query points to obtain a plurality of point attributes for the set of query points,
• processing, for each trajectory of at least a subset of trajectories, the plurality of point attributes corresponding to the set of points of interest to obtain a trajectory cost for the trajectory,
• selecting, from the set of trajectories, a selected trajectory according to trajectory cost, and
• outputting, by the virtual driver, a control signal according to the selected trajectory to control the autonomous system through the geographic region.

A non-transitory computer readable medium comprising computer readable program code for causing a computer system to perform operations comprising:

• obtaining, by a virtual driver of an autonomous system, a set of trajectories through a geographic region for the autonomous system;
• generating, for each trajectory in the set of trajectories, a set of points of interest in the geographic region to obtain sets of points of interest;
• quantizing the sets of points of interest to obtain a set of query points in the geographic region;
• querying an implicit decoder model with the set of query points to obtain a plurality of point attributes for the set of query points;
• processing, for each trajectory of at least a subset of trajectories, the plurality of point attributes corresponding to the set of points of interest to obtain a trajectory cost for the trajectory;
• selecting, from the set of trajectories, a selected trajectory according to trajectory cost;
• and
• outputting, by the virtual driver, a control signal according to the selected trajectory to control the autonomous system through the geographic region.