Map consistency checker
Techniques relating to monitoring map consistency are described. In an example, a monitoring component associated with a vehicle can receive sensor data associated with an environment in which the vehicle is positioned. The monitoring component can generate, based at least in part on the sensor data, an estimated map of the environment, wherein the estimated map is encoded with policy information for driving within the environment. The monitoring component can then compare first information associated with a stored map of the environment with second information associated with the estimated map to determine whether the estimated map and the stored map are consistent. Component(s) associated with the vehicle can then control the object based at least in part on results of the comparing.
1. One or more non-transitory computer-readable media storing instructions, that when executed by one or more processors, cause the one or more processors to perform operations comprising:
receiving sensor data;
generating, based at least in part on the sensor data, an estimated map of an environment comprising policy information for driving within the environment;
comparing first information associated with a stored map of the environment with second information associated with the estimated map to determine whether the stored map is consistent with the estimated map, the first information is a trajectory, determined based at least in part on the stored map, along which an object is to follow and the second information is a label associated with a portion of the estimated map to which the trajectory corresponds; and
controlling the object based at least in part on results of the comparing.
2. The one or more non-transitory computer-readable media of claim 1 , wherein the policy information comprises one or more of:
a first indication of a drivable surface,
a second indication of a lane boundary, or
a third indication of a direction of traffic.
3. The one or more non-transitory computer-readable media of claim 1 , the operations further comprising:
determining, based at least in part on the first information and the second information, a difference in one or more of a position or orientation of the object; and
determining whether the difference meets or exceeds a threshold difference.
4. The one or more non-transitory computer-readable media of claim 3 , wherein determining the difference comprises determining whether a first pixel of the estimated map differs from a second pixel of the stored map.
5. The one or more non-transitory computer-readable media of claim 1 , wherein the sensor data comprises one or more of image data or lidar data, the operations further comprising:
determining, based on the sensor data, a top-down representation of the environment; and
determining the estimated map based at least in part on the top-down representation.
6. The one or more non-transitory computer-readable media of claim 5 , wherein the sensor data comprises image data, and wherein determining the top-down representation of the environment comprises:
re-sizing, as resized image, images associated with the image data to a designated size;
encoding, as encoded images, the resized images using a first residual neural network;
converting, as converted images the encoded images from a projection view to an orthogonal view using a fully connected neural network; and
decoding the converted images in the orthogonal view using a second residual neural network.
7. The one or more non-transitory computer-readable media of claim 1 , the operations further comprising:
determining, based at least in part on the sensor data, a pose of the object; and
accessing a portion of the stored map of the environment for the comparing based at least in part on the pose of the object.
8. The one or more non-transitory computer-readable media of claim 1 , the operations further comprising:
determining, based at least in part on comparing the first information with the second information, a confidence score, wherein the confidence score indicates at least one of:
that a trajectory associated with the object is on a drivable surface of the estimated map;
that the trajectory will cross a lane line;
that a pose of the object is reliable; or
that information associated with the stored map is reliable; and
determining whether the stored map is consistent with the estimated map based at least in part on the confidence score.
9. The one or more non-transitory computer-readable media of claim 8 , the operations further comprising:
comparing the confidence score with a threshold; and
determining that the stored map is inconsistent with the estimated map based at least in part on the confidence score being at or below the threshold.
10. A system comprising:
one or more processors; and
one or more non-transitory computer-readable media storing instructions, that when executed by one or more processors, cause the one or more processors to perform operations comprising:
receiving sensor data;
generating, based at least in part on the sensor data, an estimated map of an environment comprising policy information for driving within the environment;
comparing first information associated with a stored map of the environment with second information associated with the estimated map to determine whether the stored map is consistent with the estimated map, the first information is a trajectory, determined based at least in part on the stored map, along which an object is to follow and the second information is a label associated with a portion of the estimated map to which the trajectory corresponds; and
controlling the object based at least in part on results of the comparing.
11. The system of claim 10 , wherein the policy information comprises one or more of:
a first indication of a drivable surface,
a second indication of a lane boundary, or
a third indication of a direction of traffic.
12. The system of claim 10 , the operations further comprising:
determining, based at least in part on the first information and the second information, a difference in one or more of a position or orientation of the object; and
determining whether the difference meets or exceeds a threshold difference.
13. The system of claim 12 , wherein determining the difference comprises determining whether a first pixel of the estimated map differs from a second pixel of the stored map.
14. The system of claim 10 , wherein the sensor data comprises one or more of image data or lidar data, the operations further comprising:
determining, based on the sensor data, a top-down representation of the environment; and
determining the estimated map based at least in part on the top-down representation.
15. The system of claim 10 , the operations further comprising:
determining, based at least in part on comparing the first information with the second information, a confidence score, wherein the confidence score indicates at least one of:
that a trajectory associated with the object is on a drivable surface of the estimated map;
that the trajectory will cross a lane line;
that a pose of the object is reliable; or
that information associated with the stored map is reliable; and
determining whether the stored map is consistent with the estimated map based at least in part on the confidence score.
16. A method comprising:
receiving sensor data;
generating, based at least in part on the sensor data, an estimated map of an environment comprising policy information for driving within the environment;
comparing first information associated with a stored map of the environment with second information associated with the estimated map to determine whether the stored map is consistent with the estimated map, the first information is a trajectory, determined based at least in part on the stored map, along which an object is to follow and the second information is a label associated with a portion of the estimated map to which the trajectory corresponds; and
controlling the object based at least in part on results of the comparing.
17. The method of claim 16 , wherein the policy information comprises one or more of:
a first indication of a drivable surface,
a second indication of a lane boundary, or
a third indication of a direction of traffic.
18. The method of claim 16 , wherein the sensor data comprises one or more of image data or lidar data, the method further comprising:
determining, based on the sensor data, a top-down representation of the environment; and
determining the estimated map based at least in part on the top-down representation.
19. The method of claim 16 , further comprising:
determining, based at least in part on comparing the first information with the second information, a confidence score, wherein the confidence score indicates at least one of:
that a trajectory associated with the object is on a drivable surface of the estimated map;
that the trajectory will cross a lane line;
that a pose of the object is reliable; or
that information associated with the stored map is reliable; and
determining whether the stored map is consistent with the estimated map based at least in part on the confidence score.
20. The method of claim 16 , further comprising:
determining, based at least in part on the first information and the second information, a difference in one or more of a position or orientation of the object; and
determining whether the difference meets or exceeds a threshold difference.