Predicting behaviors of oncoming vehicles

1. A method for world objects tracking and prediction by an autonomous vehicle (AV), comprising:

identifying, based on first observation data received from sensors of the AV, an oncoming vehicle;

identifying, based on second observation data received from the sensors of the AV, a road object;

generating a lane-following hypothesis for the oncoming vehicle, the lane-following hypothesis indicating an intention that the oncoming vehicle remain in a current road lane;

computing a lane-following reference driveline for the lane-following hypothesis of the oncoming vehicle;

in response to determining that the lane-following reference driveline is blocked by the road object,

generating a go-around hypothesis for the oncoming vehicle; and

computing a go-around reference driveline for the go-around hypothesis; and

providing at least one of a go-around trajectory corresponding to the go-around hypothesis or a lane-following trajectory corresponding to the lane-following hypothesis.

2. The method of claim 1 , further comprising:

updating the lane-following reference driveline of the lane-following hypothesis; and

in response to determining that the updated lane-following reference driveline is not blocked, removing the go-around hypothesis.

3. The method of claim 1 , further comprising:

determining a go-around likelihood that the oncoming vehicle follows the go-around hypothesis and a lane-following likelihood that the oncoming vehicle follows the lane-following hypothesis.

4. The method of claim 3 , further comprising:

in response to determining that the go-around likelihood is below a first threshold, operating the AV to continue along a current trajectory.

5. The method of claim 3 , further comprising:

in response to determining that the go-around likelihood exceeds a second threshold and the go-around reference driveline blocks a current path of the AV at a blockage location, determining at least one of a speed or a deceleration profile to bring the AV to a stop at the blockage location.

6. The method of claim 1 , wherein determining that the lane-following reference driveline is blocked by the road object comprises:

using map information to determine that the lane-following reference driveline is blocked.

7. The method of claim 6 , wherein providing the go-around trajectory comprises:

predicting the go-around trajectory using a motion model of the oncoming vehicle, such that a current predicted path of the oncoming vehicle coincides with the lane-following reference driveline.

8. The method of claim 7 , wherein the go-around trajectory is provided for a predetermined future length of time.

9. The method of claim 8 , wherein the predetermined future length of time is 6 seconds.

10. The method of claim 1 , further comprising:

in response to determining that the go-around trajectory blocks a current path of the AV at a blockage location, determining at least one of a speed or a deceleration profile to bring the AV to a stop at the blockage location.

11. A system for world objects tracking and prediction by an autonomous vehicle (AV), comprising

a memory; and

a processor, the memory includes instructions executable by the processor to:

identify, based on first observation data received from sensors of the AV, an oncoming vehicle;

identify, based on second observation data received from the sensors of the AV, a road object;

generate a lane-following hypothesis for the oncoming vehicle, the lane-following hypothesis indicating an intention that the oncoming vehicle remain in a current road lane;

compute a lane-following reference driveline for the lane-following hypothesis of the oncoming vehicle;

in response to determining that the lane-following reference driveline is blocked by the road object,

generate a go-around hypothesis for the oncoming vehicle; and

compute a go-around reference driveline for the go-around hypothesis; and

provide at least one of a go-around trajectory corresponding to the go-around hypothesis or a lane-following trajectory corresponding to the lane-following hypothesis.

12. The system of claim 11 , wherein instructions further comprise instructions to:

update the lane-following reference driveline of the lane-following hypothesis; and

in response to determining that the updated lane-following reference driveline is not blocked, remove the go-around hypothesis.

13. The system of claim 11 , wherein instructions further comprise instructions to:

determine a go-around likelihood that the oncoming vehicle follows the go-around hypothesis and a lane-following likelihood that the oncoming vehicle follows the lane-following hypothesis.

14. The system of claim 13 , wherein instructions further comprise instructions to:

in response to determining that the go-around likelihood is below a first threshold, operate the AV to continue along a current trajectory.

15. The system of claim 13 , wherein instructions further comprise instructions to:

in response to determining that the go-around likelihood exceeds a second threshold and the go-around reference driveline blocks a current path of the AV at a blockage location, determine at least one of a speed or a deceleration profile to bring the AV to a stop at the blockage location.

16. The system of claim 11 , wherein to determine that the lane-following reference driveline is blocked by the road object comprises to:

use map information to determine that the lane-following reference driveline is blocked.

17. The system of claim 16 , wherein to providing the go-around trajectory comprises to:

predict the go-around trajectory using a motion model of the oncoming vehicle, such that a current predicted path of the oncoming vehicle coincides with the lane-following reference driveline.

18. The system of claim 17 , wherein the go-around trajectory is provided for a predetermined future length of time.

19. The system of claim 18 , wherein the predetermined future length of time is 6 seconds.

20. The system of claim 11 , wherein instructions further comprise instructions to:

in response to determining that the go-around trajectory blocks a current path of the AV at a blockage location, determine at least one of a speed or a deceleration profile to bring the AV to a stop at the blockage location.