Determining perceptual spatial relevancy of objects and road actors for automated driving
Disclosed herein are system, method, and computer program product embodiments for determining objects that are kinematically capable, even if non-compliant with rules-of-the-road, of affecting a trajectory of a vehicle. The computing system (e.g., perception system, etc.) of a vehicle may generate a trajectory for the vehicle and a respective trajectory for each object of a plurality of objects within a field of view (FOV) of the sensing device associated with the vehicle. The computing system may identify objects of the plurality of objects with trajectories that intersect the trajectory for the vehicle and remove from such objects, objects with trajectories that at least one of exit the FOV or intersect with other objects of the plurality of objects within the FOV. The computing system may select, from remaining objects with trajectories that intersect the trajectory for the vehicle, objects with trajectories that indicate a respective collision between the object and the vehicle and assign a severity of the respective collision.
1. A computer-implemented method comprising:
generating, by one or more computing devices of a vehicle, based on sensor information received from a sensing device associated with the vehicle, a trajectory for the vehicle and a respective trajectory for each object of a plurality of objects within a field of view (FOV) of the sensing device;
identifying, based on the respective trajectories for each object of the plurality of objects, objects of the plurality of objects with trajectories that intersect the trajectory for the vehicle;
removing, from the objects with trajectories that intersect the trajectory for the vehicle, objects with trajectories that intersect with other objects of the plurality of objects within the FOV;
selecting, from remaining objects with trajectories that intersect the trajectory for the vehicle, objects with trajectories that indicate a respective collision between the object and the vehicle;
assigning, for each object of the objects with the trajectories that indicate the respective collision between the object and the vehicle, a severity of the respective collision; and
causing, based on the severity of the respective collision for at least one object of the objects with the trajectories that indicate the respective collision, the vehicle to perform a driving maneuver.
2. The method of claim 1 , further comprising detecting, based on the sensor information, each object of the plurality of objects within the FOV,
wherein each object of the plurality of objects within the FOV satisfies a perception threshold that indicates an amount of an object sensed by the sensing device that is occluded by an item.
3. The method of claim 1 , wherein, for the vehicle and each object of the plurality of objects, the sensor information indicates a respective position and a respective velocity, and wherein the generating the trajectory for the vehicle and the respective trajectories for each object of the plurality of objects is based on the respective positions and the respective velocities.
4. The method of claim 1 , wherein the assigning, for each object of the objects with the trajectories that indicate the respective collision, the severity of the respective collision comprises:
inputting, into a predictive model, a velocity of the vehicle, a position of the vehicle, a velocity for the object, and an object type for the object; and
receiving from the predictive model, based on the velocity of the vehicle, the position of the vehicle, the velocity for the object, and the object type, an indication of the severity of the respective collision.
5. The method of claim 1 , further comprising causing, for each object of the objects with the trajectories that indicate the respective collision between the object and the vehicle, display of the respective trajectory and the trajectory for the vehicle.
6. The method of claim 1 , further comprising sending, to a user device, an indication of the severity of the respective collision for each object of the objects with the trajectories that indicate the respective collision between the object and the vehicle.
7. A computing system comprising:
a memory of a vehicle configured to store instructions;
a processor of the vehicle, coupled to the memory, configured to process the stored instructions to:
generate, based on sensor information received from a sensing device associated with a vehicle, a trajectory for the vehicle and a respective trajectory for each object of a plurality of objects within a field of view (FOV) of the sensing device;
identify, based on the respective trajectories for each object of the plurality of objects, objects of the plurality of objects with trajectories that intersect the trajectory for the vehicle;
remove, from the objects with trajectories that intersect the trajectory for the vehicle, objects with trajectories that intersect with other objects of the plurality of objects within the FOV;
select, from remaining objects with trajectories that intersect the trajectory for the vehicle, objects with trajectories that indicate a respective collision between the object and the vehicle;
assign, for each object of the objects with the trajectories that indicate the respective collision between the object and the vehicle, a severity of the respective collision; and
cause, based on the severity of the respective collision for at least one object of the objects with the trajectories that indicate the respective collision, the vehicle to perform a driving maneuver.
8. The system of claim 7 , the processor further configured to detect, based on the sensor information, each object of the plurality of objects within the FOV,
wherein each object of the plurality of objects within the FOV satisfies a perception threshold that indicates an amount of an object sensed by the sensing device that is occluded by an item.
9. The system of claim 7 , wherein, for the vehicle and each object of the plurality of objects, the sensor information indicates a respective position and a respective velocity, and the processor configured to generate the trajectory for the vehicle and the respective trajectories for each object of the plurality of objects is further configured to generate the trajectory for the vehicle and the respective trajectories for each object of the plurality of objects based on the respective positions and the respective velocities.
10. The system of claim 7 , the processor configured to assign, for each object of the objects with the trajectories that indicate the respective collision, the severity of the respective collision is further configured to:
input, into a predictive model, a velocity of the vehicle, a position of the vehicle, a velocity for the object, and an object type for the object; and
receive from the predictive model, based on the velocity of the vehicle, the position of the vehicle, the velocity for the object, and the object type, an indication of the severity of the respective collision.
11. The system of claim 7 , the processor further configured to cause, for each object of the objects with the trajectories that indicate the respective collision between the object and the vehicle, display of the respective trajectory and the trajectory for the vehicle.
12. The system of claim 7 , the processor further configured to send, to a user device, an indication of the severity of the respective collision for each object of the objects with the trajectories that indicate the respective collision between the object and the vehicle.
13. A non-transitory computer-readable medium having instructions stored thereon that, when executed by at least one computing device, cause the at least one computing device to perform operations comprising:
generating, based on sensor information received from a sensing device associated with a vehicle, a trajectory for the vehicle and a respective trajectory for each object of a plurality of objects within a field of view (FOV) of the sensing device;
identifying, based on the respective trajectories for each object of the plurality of objects, objects of the plurality of objects with trajectories that intersect the trajectory for the vehicle;
removing, from the objects with trajectories that intersect the trajectory for the vehicle, objects with trajectories that intersect with other objects of the plurality of objects within the FOV;
selecting, from remaining objects with trajectories that intersect the trajectory for the vehicle, objects with trajectories that indicate a respective collision between the object and the vehicle;
assigning, for each object of the objects with the trajectories that indicate the respective collision between the object and the vehicle, a severity of the respective collision; and
causing, based on the severity of the respective collision for at least one object of the objects with the trajectories that indicate the respective collision, the vehicle to perform a driving maneuver.
14. The non-transitory computer-readable medium claim 13 , further comprising detecting, based on the sensor information, each object of the plurality of objects within the FOV,
wherein each object of the plurality of objects within the FOV satisfies a perception threshold that indicates an amount of an object sensed by the sensing device that is occluded by an item.
15. The non-transitory computer-readable medium claim 13 , wherein, for the vehicle and each object of the plurality of objects, the sensor information indicates a respective position and a respective velocity, and wherein the generating the trajectory for the vehicle and the respective trajectories for each object of the plurality of objects is based on the respective positions and the respective velocities.
16. The non-transitory computer-readable medium claim 13 , wherein the assigning, for each object of the objects with the trajectories that indicate the respective collision, the severity of the respective collision comprises:
inputting, into a predictive model, a velocity of the vehicle, a position of the vehicle, a velocity for the object, and an object type for the object; and
receiving from the predictive model, based on the velocity of the vehicle, the position of the vehicle, the velocity for the object, and the object type, an indication of the severity of the respective collision.
17. The non-transitory computer-readable medium claim 13 , further comprising causing, for each object of the objects with the trajectories that indicate the respective collision between the object and the vehicle, display of the respective trajectory and the trajectory for the vehicle.
18. The non-transitory computer-readable medium claim 13 , further comprising sending, to a user device, an indication of the severity of the respective collision for each object of the objects with the trajectories that indicate the respective collision between the object and the vehicle.