Control of autonomous vehicle based on determined yaw parameter(s) of additional vehicle
Determining an instantaneous vehicle characteristic (e.g., at least one yaw rate) of an additional vehicle that is in addition to a vehicle being autonomously controlled, and adapting autonomous control of the vehicle based on the determined instantaneous vehicle characteristic of the additional vehicle. For example, autonomous steering, acceleration, and/or deceleration of the vehicle can be adapted based on a determined instantaneous vehicle characteristic of the additional vehicle. In many implementations, the instantaneous vehicle characteristics of the additional vehicle are determined based on data from a phase coherent Light Detection and Ranging (LIDAR) component of the vehicle, such as a phase coherent LIDAR monopulse component and/or a frequency-modulated continuous wave (FMCW) LIDAR component.
1. A computer implemented method, comprising:
receiving, from a phase coherent Light Detection and Ranging (LIDAR) component of a vehicle, a group of LIDAR data points of a sensing cycle of the phase coherent LIDAR component;
each of the LIDAR data points of the group indicating a corresponding instantaneous range and a corresponding instantaneous velocity for a corresponding point in an environment of the vehicle;
each of the LIDAR data points of the group being generated based on a corresponding sensing event of the phase coherent LIDAR component during the sensing cycle; and
processing the group of LIDAR data points with a machine learning model to generate an output that indicates that a subset of the group of LIDAR data points correspond to a dynamic object in the environment; and
controlling an autonomous vehicle based on a determined at least one instantaneous characteristic of the dynamic object.
2. The computer implemented method of claim 1 , wherein the dynamic object is at least one of: a pedestrian and an additional vehicle.
3. The computer implemented method of claim 1 , wherein processing of the group of LIDAR data points with the machine learning model further comprises determining an instantaneous characteristic of the dynamic object.
4. The computer implemented method of claim 3 , wherein determining the instantaneous characteristic of the dynamic object comprises:
determining a first set of one or more of a plurality of the LIDAR data points of the group;
determining a second set of one or more of the plurality of the LIDAR data points of the group based on the second set being spatially offset from the first set; and
determining the at least one instantaneous characteristic based on comparison of one or more of the corresponding instantaneous velocities of the first set to one or more of the corresponding instantaneous velocities of the second set.
5. The computer implemented method of claim 1 , further comprising determining at least one instantaneous characteristic of the dynamic object based on the corresponding instantaneous velocity of the LIDAR data points of the group and based on a stored model for the dynamic object, the stored model describing geometric features of the dynamic object.
6. The computer implemented method of claim 5 , wherein the stored model is a three-dimensional model of the dynamic object.
7. The computer implemented method of claim 6 , further comprising selecting the stored model, from multiple candidate stored models, based on determining that sensor data corresponds most closely to the stored model.
8. The computer implemented method of claim 7 , wherein the sensor data comprises the group of LIDAR data points.
9. The computer implemented method of claim 5 , wherein determining the at least one instantaneous characteristic, further comprises:
determining a velocity differential based on comparison of a first set velocity magnitude and a second set velocity magnitude, wherein the first set velocity magnitude is based on a first set of the corresponding instantaneous velocities of the LIDAR data points of the group, and wherein the second set velocity magnitude is based on a second set of the corresponding instantaneous velocities of the LIDAR data points of the group; and
converting the velocity differential to the at least one instantaneous characteristic based on the stored model for the dynamic object.
10. The computer implemented method of claim 1 , wherein the at least one instantaneous characteristic of the dynamic object is a velocity differential that is indicative of yaw rate, and wherein adapting control of the vehicle based on the determined at least one instantaneous characteristic of the dynamic object comprises adapting a velocity of the vehicle and/or a direction of the vehicle based on the velocity differential exceeding a threshold.
11. The computer implemented method of claim 1 , wherein the controlling the autonomous vehicle based on the determined at least one instantaneous characteristic of the dynamic object comprises:
determining at least one candidate trajectory of the dynamic object based on the determined at least one instantaneous characteristic; and
adapting autonomous control of the vehicle based on the at least one candidate trajectory.
12. The computer implemented method of claim 11 , wherein the dynamic object is an additional vehicle and wherein the controlling the autonomous vehicle based on the at least one candidate trajectory comprises performing autonomous evasive steering.
13. The computer implemented method of claim 1 , wherein the LIDAR component is a LIDAR monopulse component and wherein corresponding sensing events of the LIDAR component each comprise a first receiver sensing event at a first coherent receiver of the LIDAR monopulse component and a second receiver sensing event at a second coherent receiver of the LIDAR monopulse component, the first coherent receiver being positionally offset from the second coherent receiver.
14. The computer implemented method of claim 13 , wherein the LIDAR data points of the group are super-resolution LIDAR data points generated based on combining the first receiver sensing events and the second receiver sensing event.
15. A computer implemented method using a phase coherent LIDAR component, comprising:
receiving, from a phase coherent Light Detection and Ranging (LIDAR) component of an autonomous vehicle, a group of LIDAR data points of a sensing cycle of the phase coherent LIDAR component;
each of the LIDAR data points of the group indicating a corresponding instantaneous range and a corresponding instantaneous velocity for a corresponding point in an environment of the autonomous vehicle;
each of the LIDAR data points of the group being generated based on a corresponding sensing event of the phase coherent LIDAR component during the sensing cycle;
processing the group of LIDAR data points with a machine learning model to generate an output that indicates that a subset of the group of LIDAR data points correspond to a dynamic object in the environment; and
providing the output of the machine learning model to one or both of: a planning subsystem of the autonomous vehicle and a control subsystem of the autonomous vehicle to control an autonomous vehicle based at least in part on the output of the machine learning model.
16. The computer implemented method of claim 15 further comprising the sensing cycle having a first receiver sensor and a second receiver sensor and combining first receiver sensing events and second receiver sensing events using one or both of: a frequency domain technique and a spatial domain technique.
17. The computer implemented method of claim 15 wherein the dynamic object is at least of one an additional vehicle and/or a pedestrian.
18. The computer implemented method of claim 15 further comprising modifying control characteristics of the autonomous vehicle based on the output of the machine learning model.
19. An autonomous vehicle, comprising:
a phase coherent Light Detection and Ranging (LIDAR) component; one or more processors executing stored computer instructions to:
receive, from the LIDAR component of the autonomous vehicle, a group of LIDAR data points of a sensing cycle of the phase coherent LIDAR component;
each of the LIDAR data points of the group indicating a corresponding instantaneous range and a corresponding instantaneous velocity for a corresponding point in an environment of the autonomous vehicle;
each of the LIDAR data points of the group being generated based on a corresponding sensing event of the phase coherent LIDAR component during the sensing cycle;
processing the group of LIDAR data points with a machine learning model to generate an output that indicates that a subset of the group of LIDAR data points correspond to a dynamic object in the environment; and
control the autonomous vehicle using output of the machine learning model.