The current dataset does not have the lane markings or traffic lights labeled as any class. I need to segment lane markings and traffic lights and hence need to be able to train on the lights and markings present in the original CamVid dataset.
How do I calculate the class weights for lane markings and traffic lights.
The current dataset does not have the lane markings or traffic lights labeled as any class. I need to segment lane markings and traffic lights and hence need to be able to train on the lights and markings present in the original CamVid dataset. How do I calculate the class weights for lane markings and traffic lights.