Method of and system for predicting a maneuver of an object
Methods and devices for generating data for controlling a Self-Driving Car (SDC) are disclosed. The method includes: receiving a section of a road map corresponding to surroundings of the SDC and at least one object, generating predicted trajectories including potential future location points of the at least one object, mapping the potential future location points on the section of the road map, computing a score for each of the potential future location points, computing an aggregated score from the scores corresponding to the potential future location points, and based on the aggregated score, determining a predicted location of the at least one object at the future instance of time.
1. A method of predicting a maneuver of at least one object, the at least one object being in a vicinity of a Self-Driving Car (SDC), the SDC being communicatively coupled to an electronic device, the method being executable by the electronic device, the method comprising:
receiving, by the electronic device,
a section of a road map corresponding to surroundings of the SDC and the at least one object, the section of the road map including information about a plurality of road lanes;
generating, by the electronic device,
a plurality of predicted trajectories associated with the at least one object, each of the plurality of predicted trajectories including potential future location points of the at least one object on the section of the road map, the plurality of predicted trajectories being generated without consideration of the plurality of road lanes on the section of the road map;
for a given one of the plurality of predicted trajectories,
mapping, by the electronic device, the potential future location points associated with the given one of the plurality of predicted trajectories on the section of the road map in accordance with the information about the plurality of road lanes;
computing, by the electronic device, scores for each of the potential future location points associated with the given one of the plurality of predicted trajectories, the scores representing an association of a given potential future location point of the potential future location points with the plurality of road lanes at a future instance of time, wherein computing the scores comprises assigning a first predetermined score to the given potential future location point if the given potential future location point and a previous potential future location point are located on a same road lane of the plurality of road lanes and assigning a second predetermined score to the given potential future location point if the given potential future location point and the previous potential future location point are not located on the same road lane of the plurality of road lanes;
computing, by the electronic device, an aggregated score from the scores corresponding to the given potential future location point of the potential future location points associated with the given one of the plurality of predicted trajectories; and
based on the aggregated score, determining, by the electronic device, a predicted location within the plurality of road lanes of the at least one object at the future instance of time, and
based on the predicted location of the at least one object:
changing an acceleration of the SDC; or
changing a speed of the SDC; or
changing a direction of the SDC to change a road lane of the plurality of road lanes in which the SDC is currently moving; or
changing the direction of the SDC to take a turn.
2. The method of claim 1 , wherein the information about the plurality of road lanes includes one or more of:
a number of the plurality of road lanes in the section of the road map;
geometrical properties of the plurality of road lanes,
the geometrical properties of the plurality of road lanes includes one or more of:
a width of each of the plurality of road lanes,
a curvature of each of the plurality of road lanes;
a number of exits on the section of the road map;
a marking on the section of the road map indicating one or more of:
a first indication of turning left or right,
a second indication of an exit,
a partitioning between the plurality of road lanes.
3. The method of claim 1 , wherein the section of the road map is a high definition (HD) map.
4. The method of claim 1 , wherein the plurality of predicted trajectories are based on movement data of the at least one object.
5. The method of claim 1 , wherein the mapping of the potential future location points on the section of the road map comprises mapping of the potential future location points associated with the given one of the plurality of predicted trajectories onto the plurality of road lanes.
6. The method of claim 1 , wherein the mapping of the potential future location points on the section of the road map comprises generating a plurality of hypotheses, wherein a given hypothesis of the plurality of hypotheses represents a relationship of the given potential future location point of the potential future location points with the plurality of road lanes.
7. The method of claim 6 , wherein the aggregated score is computed as one or more of:
a summation of the scores assigned to the given potential future location point associated with the given one of the plurality of predicted trajectories in accordance with the plurality of hypotheses;
an average of the scores assigned to the given potential future location point associated with the given one of the plurality of predicted trajectories in accordance with the plurality of hypotheses.
8. The method of claim 1 , wherein the first predetermined score is higher than the second predetermined score.
9. The method of claim 1 , wherein the SDC and the at least one object are moving in at least one of:
a same direction;
an opposite direction.
10. The method of claim 1 , wherein the computing the scores for each of the potential future location points comprises:
using a machine learning algorithm (MLA), in which the potential future location points are used as input, and
an output of the MLA are the scores corresponding to the given potential future location point of the potential future location points.
11. The method of claim 1 , wherein
a higher aggregated score associated with the given potential future location point is representative of a fact that the at least one object is unlikely to make rapid lane changes between sequential periods of time with a small time change increment therebetween;
a lower aggregated score associated with the given potential future location point is representative of unlikely zig-zagging behavior of the at least one object between sequential periods of time with the small time change increment therebetween.
12. The method of claim 1 , wherein the object is a dynamic object.
13. The method of claim 1 , wherein the predicted location is one of the potential future location points.
14. The method of claim 1 , wherein the predicted location is being located at one or more of:
an end of the given one of the plurality of predicted trajectories;
a beginning of the given one of the plurality of predicted trajectories;
a first future location point associated with the given one of the plurality of predicted trajectories located on a road lane of the plurality of road lanes in which the SDC is moving.
15. An electronic device for predicting a maneuver of at least one object, the at least one object being in a vicinity of a Self-Driving Car (SDC), the electronic device communicatively coupled to the SDC, the electronic device configured to:
receive a section of a road map corresponding to surroundings of the SDC and the at least one object, the section of the road map including information about a plurality of road lanes;
generate, a plurality of predicted trajectories associated with the at least one object, each of the plurality of predicted trajectories including potential future location points of the at least one object on the section of the road map, the plurality of predicted trajectories being generated without consideration of the plurality of road lanes on the section of the road map;
for a given one of the plurality of predicted trajectories,
map the potential future location points associated with the given one of the plurality of predicted trajectories on the section of the road map in accordance with the information about the plurality of road lanes;
compute scores for each of the potential future location points associated with the given one of the plurality of predicted trajectories, the scores representing an association of a given potential future location point of the potential future location points with the plurality of road lanes at a future instance of time, wherein computing the scores comprises assigning a first predetermined score to the given potential future location point if the given potential future location point and a previous potential future location point are located on a same road lane of the plurality of road lanes and assigning a second predetermined score to the given potential future location point if the given potential future location point and the previous potential future location point are not located on the same road lane of the plurality of road lanes;
compute an aggregated score from the scores corresponding to the given potential future location point of the potential future location points associated with the given one of the plurality of predicted trajectories; and
based on the aggregated score, determine a predicted location within the plurality of road lanes of the at least one object at the future instance of time, and
based on the predicted location of the at least one object:
instruct the SDC to change an acceleration of the SDC; or
instruct the SDC to change a speed of the SDC; or
instruct the SDC to change a direction of the SDC to change a road lane of the plurality of road lanes in which the SDC is currently moving; or
instruct the SDC to change the direction of the SDC to take a turn.
16. The electronic device of claim 15 , wherein the information about the plurality of road lanes includes one or more of:
a number of the plurality of road lanes in the section of the road map;
geometrical properties of the plurality of road lanes,
the geometrical properties of the plurality of road lanes includes one or more of:
a width of each of the plurality of road lanes,
a curvature of each of the plurality of road lanes;
a number of exits on the section of the road map;
a marking on the section of the road map indicating one or more of:
a first indication of turning left or right,
a second indication of an exit,
a partitioning between the plurality of road lanes.
17. The electronic device of claim 15 , wherein the section of the road map is a high definition (HD) map.
18. A method of predicting a maneuver of at least one object, the at least one object being in a vicinity of a Self-Driving Car (SDC), the SDC being communicatively coupled to an electronic device, the method being executable by the electronic device, the method comprising:
receiving, by the electronic device,
a section of a road map corresponding to surroundings of the SDC and the at least one object, the section of the road map including information about a plurality of road lanes;
generating, by the electronic device,
a plurality of predicted trajectories associated with the at least one object, each of the plurality of predicted trajectories including potential future location points of the at least one object on the section of the road map, the plurality of predicted trajectories being generated without consideration of the plurality of road lanes on the section of the road map;
for a given one of the plurality of predicted trajectories,
mapping, by the electronic device, the potential future location points associated with the given one of the plurality of predicted trajectories on the section of the road map in accordance with the information about the plurality of road lanes;
computing, by the electronic device, scores for each of the potential future location points associated with the given one of the plurality of predicted trajectories, the scores representing an association of a given potential future location point of the potential future location points with the plurality of road lanes at a future instance of time, wherein computing the scores comprises assigning a first predetermined score to the given potential future location point if the given potential future location point and a next potential future location point are located on a same road lane of the plurality of road lanes and assigning a second predetermined score to the given potential future location point if the given potential future location point and the next potential future location point are not located on the same road lane of the plurality of road lanes;
computing, by the electronic device, an aggregated score from the scores corresponding to the given potential future location point of the potential future location points associated with the given one of the plurality of predicted trajectories; and
based on the aggregated score, determining, by the electronic device, a predicted location within the plurality of road lanes of the at least one object at the future instance of time, and
based on the predicted location of the at least one object:
changing an acceleration of the SDC; or
changing a speed of the SDC; or
changing a direction of the SDC to change a road lane of the plurality of road lanes in which the SDC is currently moving; or
changing the direction of the SDC to take a turn.