Devices and methods for accurately identifying objects in a vehicle's environment
AssigneeIntel Corporation
InventorsNilesh Ahuja, Ibrahima Ndiour, Javier Felip Leon, David Gomez Gutierrez, Ranganath Krishnan, Mahesh Subedar, Omesh Tickoo
Life of this patent
Estimating the term…
Abstract
Vehicle navigation control systems in autonomous driving rely on accurate predictions of objects within the vicinity of the vehicle to appropriately control the vehicle safely through its surrounding environment. Accordingly this disclosure provides methods and devices which implement mechanisms for obtaining contextual variables of the vehicle's environment for use in determining the accuracy of predictions of objects within the vehicle's environment.
Claims18
A perception device for a vehicle to determine an uncertainty metric of an object, the perception device comprising one or more processors configured to:
- detect a vehicle environment;
- obtain a plurality of contextual variables from a plurality of auxiliary devices;
- wherein the plurality of contextual variables are of interrelated conditions in which the vehicle exists;
- determine a plurality of objects within the vehicle environment;
- determine the uncertainty metric of the at least one of the plurality of objects based on a probability that the at least one of the plurality of objects occurs given at least one of the plurality of contextual variables;
- and
- implement a probabilistic model of a relationship between the plurality of objects and the plurality of contextual variables.
A method to determine an uncertainty metric of an object, the method comprising:
- detecting a vehicle environment;
- obtaining a plurality of contextual variables from a plurality of auxiliary devices, wherein the plurality of contextual variables are of interrelated conditions in which the vehicle exists;
- determining a plurality of objects within the vehicle environment;
- and
- determining the uncertainty metric of the at least one of the plurality of objects based on a probability that the at least one of the plurality of objects occurs given at least one of the plurality of contextual variables, wherein determining the uncertainty metric comprises using a probabilistic model of a relationship between the plurality of objects and the plurality of contextual variables.
A device for a vehicle to incorporate annotated data, the device comprising one or more processors configured to:
- obtain an uncertainty metric of an object, wherein the uncertainty metric describes a probability that the object occurs;
- determine an uncertainty metric threshold based on a previously trained model of the continuous learning device;
- compare the uncertainty metric with the uncertainty metric threshold;
- determine an annotation method based on the comparison of the uncertainty metric and the uncertainty metric threshold;
- collect data associated with the object;
- annotate, according to the determined annotation method, the data to more accurately identify the object;
- and
- retrain the model of the continuous learning device with the annotated data.
Dependencies are inferred from the claim text.