Closed Ace-Xu closed 4 years ago
I trained with CULane dataset and when I use the CULane evaluation code, this error occured: lane size must be greater or equal to 2 Can you help me?
I think during your prediction for a lane, your model predicts an area which only contains one evaluation dots. You can add post-processing code to filter out this kind of situation.
I trained with CULane dataset and when I use the CULane evaluation code, this error occured: lane size must be greater or equal to 2 Can you help me?