System and method for updating an autonomous vehicle driving model based on the vehicle driving model becoming statistically incorrect
Systems and methods for implementing one or more autonomous features for autonomous and semi-autonomous control of one or more vehicles are provided. More specifically, image data may be obtained from an image acquisition device and processed utilizing one or more machine learning models to identify, track, and extract one or more features of the image utilized in decision making processes for providing steering angle and/or acceleration/deceleration input to one or more vehicle controllers. In some instances, techniques may be employed such that the autonomous and semi-autonomous control of a vehicle may change between vehicle follow and lane follow modes. In some instances, at least a portion of the machine learning model may be updated based on one or more conditions.
1. A method comprising:
autonomously driving a path generated by a first autonomous vehicle model;
receiving a statistical accuracy associated with the first autonomous vehicle model based on comparison of the path generated by the first autonomous vehicle model and a second path followed by an autonomous vehicle;
comparing the statistical accuracy to a threshold;
based on the comparison, determining that the statistical accuracy associated with the first autonomous vehicle model indicates that the first autonomous vehicle model is statistically incorrect;
updating the first autonomous vehicle model to a second autonomous vehicle model based on the determination that the first autonomous vehicle model is statistically incorrect;
autonomously driving a third path generated by the second autonomous vehicle model; and
wherein determining that the statistical accuracy associated with the first autonomous vehicle model indicates that the first autonomous vehicle model is statistically incorrect includes determining that the statistical accuracy associated with the first autonomous vehicle model is less than a threshold.
2. The method of claim 1 , further comprising:
receiving at least one update for the first autonomous vehicle model;
applying the at least one update to the first autonomous vehicle model; and
generating the second autonomous vehicle model based on the at least one update.
3. The method of claim 2 , wherein the at least one update includes one or more model parameters for a portion of the first autonomous vehicle model.
4. The method of claim 1 , wherein the statistical accuracy associated with the first autonomous vehicle model is based on at least one of a quantity of course corrections or a quantity of course deviations.
5. The method of claim 4 , wherein a course correction includes determining that an input associated with a manual override was received.
6. The method of claim 4 , wherein a course deviation includes determining that the path followed by a path to be traveled by an the autonomous vehicle is different from a projected the generated path to be traveled by the autonomous vehicle.
7. The method of claim 1 , further comprising: providing the second autonomous vehicle model to the autonomous vehicle.
8. The method of claim 7 , further comprising: providing the second autonomous vehicle model to a second autonomous vehicle.
9. The method of claim 1 , wherein the second autonomous vehicle model is generated at the autonomous vehicle.
10. A system comprising:
a memory;
a processor in communication with the memory and with an autonomous vehicle autonomously driving a path generated by a first autonomous vehicle model, wherein the processor executes instructions stored in the memory, which cause the processor to execute a method, the method comprising:
from the autonomous vehicle, receiving a statistical accuracy associated with the first autonomous vehicle model, based on comparison of the path generated by the first autonomous vehicle model and a second path followed by the autonomous vehicle;
comparing the statistical accuracy to a threshold;
based on the comparison, determining that the statistical accuracy associated with the first autonomous vehicle model indicates that the first autonomous vehicle model is statistically incorrect;
updating the first autonomous vehicle model to a second autonomous vehicle model based on the determination that the first autonomous vehicle model is statistically incorrect;
sending the second autonomous vehicle model to the autonomous vehicle, wherein the autonomous vehicle autonomously drives a third path generated by the second autonomous vehicle model; and
wherein determining that the statistical accuracy associated with the first autonomous vehicle model indicates that the first autonomous vehicle model is statistically incorrect includes determining that the statistical accuracy associated with the first autonomous vehicle model is less than a threshold.
11. The system of claim 10 , wherein the method includes: receiving at least one update for the first autonomous vehicle model; applying the at least one update to the first autonomous vehicle model; and
generating the second autonomous vehicle model based on the at least one update.
12. The system of claim 11 , wherein the at least one update includes one or more model parameters for a portion of the first autonomous vehicle model.
13. The system of claim 10 , wherein the statistical accuracy associated with the first autonomous vehicle model is based on at least one of a quantity of course corrections or a quantity of course deviations.
14. The system of claim 13 , wherein a course correction includes determining that an input associated with a manual override was received.
15. A non-transitory computer readable medium having stored thereon instructions, which when executed by a processor cause the processor to execute a method, the method comprising:
from an autonomous vehicle autonomously driving a path generated by a first autonomous vehicle model, receiving a statistical accuracy associated with the first autonomous vehicle model, based on comparison of the path generated by the first autonomous vehicle model and a second path followed by the autonomous vehicle;
comparing the statistical accuracy to a threshold;
based on the comparison, determining that the statistical accuracy associated with the first autonomous vehicle model indicates that the first autonomous vehicle model is statistically incorrect;
updating the first autonomous vehicle model to a second autonomous vehicle model based on the determination that the first autonomous vehicle model is statistically incorrect;
sending the second autonomous vehicle model to the autonomous vehicle, wherein the autonomous vehicle autonomously drives a third path generated by the second autonomous vehicle model; and
wherein determining that the statistical accuracy associated with the first autonomous vehicle model indicates that the first autonomous vehicle model is statistically incorrect includes determining that the statistical accuracy associated with the first autonomous vehicle model is less than a threshold.
16. The non-transitory computer readable medium of claim 15 , wherein the method includes:
receiving at least one update for the first autonomous vehicle model; applying the at least one update to the first autonomous vehicle model; and generating the second autonomous vehicle model based on the at least one update.
17. The non-transitory computer readable medium of claim 16 , wherein the at least one update includes one or more model parameters for a portion of the first autonomous vehicle model.
18. The non-transitory computer readable medium of claim 15 , wherein the statistical accuracy associated with the first autonomous vehicle model is based on at least one of a quantity of course corrections or a quantity of course deviations.
19. The non-transitory computer readable medium of claim 18 , wherein a course correction includes determining that an input associated with a manual override was received.