System and method for road sign ground truth construction with a knowledge graph and machine learning
A method of road sign classification utilizing a knowledge graph, including detecting and selecting a representation of a sign across a plurality of frames, outputting a prompt initiating a request for a classification associated with the representation of the sign, classifying one or more images including the sign, querying the knowledge graph to obtain a plurality of road sign classes with at least one same attribute as the sign, and classifying the sign across the plurality of frames in response to a confidence level exceeding a threshold.
1. A system comprising:
an input interface configured to receive one or more images;
a controller in communication with the input interface and configured to:
detect and select a road sign identified across a plurality of frames from the one or more images;
output a prompt initiating a request for a classification of the road sign;
detect and classify one or more images including the road signs utilizing a machine learning model;
query a knowledge graph to obtain a plurality of road sign classes with a same attribute as candidate classes for a next classifier;
classify the road sign across the plurality of frames; and
track the road sign across the plurality of frames.
2. The system of claim 1 , wherein the controller is further configured to identify one or more attributes associated with the road sign and output a plurality of road sign templates in response to the one or more attributes.
3. The system of claim 1 , wherein the controller is further configured to classify the road sign in response to input received at an interface.
4. The system of claim 1 , wherein the controller is further configured to output a first classification and, in response to a wrong classification of the first classification, receive input to re-classification from an annotator.
5. The system of claim 1 , wherein the controller is further configured to classify the road sign across the plurality of frames in response to a confidence level exceeding a threshold and output a request for attribute input when the confidence level is below the threshold.
6. The system of claim 1 , wherein the controller is further configured to output a classification associated with the road sign at a user interface.
7. A method of road sign classification utilizing a knowledge graph, comprising:
detecting and selecting a representation of a sign across a plurality of frames;
outputting a prompt initiating a request for a classification associated with the representation of the sign;
classifying one or more images including the sign;
querying the knowledge graph to obtain a plurality of road sign classes with at least one same attribute as the sign; and
classifying the sign across the plurality of frames in response to a confidence level exceeding a threshold.
8. The method of claim 7 , wherein the classification assigns an identification associated with the one or more images including signs.
9. The method of claim 7 , wherein the classification utilizes one or more machine learning models for detection of road signs, classification of road signs, and prediction of road sign properties.
10. The method of claim 7 , wherein the method includes utilizing metric learning to separate different classes in an embedded space.
11. The method of claim 7 , wherein the method includes outputting a request for attribute input associated with the representation of the sign when the confidence level is below the threshold.
12. The method of claim 11 , wherein the method includes outputting a plurality of potential sign templates in response to the attribute input.
13. A system for road sign classification utilizing a machine learning model, comprising:
a display configured to output a user interface; and
a processor in communication with the display, the processor programmed to:
detect and select a representation of a sign across one or more images utilizing the machine learning model;
output a prompt at the user interface initiating a request for a classification associated with the representation of the sign;
classify one or more images including the sign;
obtain a plurality of road sign classes associated with candidates including at least one same attribute as the sign; and
classify the sign across the one or more images in response to a confidence level exceeding a threshold.
14. The system of claim 13 , wherein the processor is further programmed to output a request for attribute input associated with the representation of the sign when the confidence level is below the threshold.
15. The system of claim 14 , wherein the processor is further programmed to output a plurality of sign templates in response to the attribute input.
16. The system of claim 13 , wherein the processor is further programmed to output one or more attributes associated with the sign.
17. The system of claim 13 , wherein the processor is further programmed to receive as input a geometric shape associated with the sign.
18. The system of claim 13 , wherein the processor is further programmed to classify utilizing a one shot classifier or a few shot classifier.
19. The system of claim 13 , wherein the processor is further programmed to obtain a plurality of road sign classes with at least one same attribute as the sign.
20. The system of claim 13 , wherein the processor is further programmed to classify utilizing a few shot classifier.