Closed wcshen closed 3 years ago
Yes that's correct. As described in the paper, the road data is not manually annotated and therefore only used for training. Further the augmentation is applied on the road data to obtain rain and fog in road scenarios with ground truth information.
Yes that's correct. As described in the paper, the road data is not manually annotated and therefore only used for training. Further the augmentation is applied on the road data to obtain rain and fog in road scenarios with ground truth information.
Hi,
So for all road data "train_road_01" and "train_road_02", we have to implement the data augmentation described in your paper ?
hi, I have downloaded the
train_road_02
,but the labels are all0
.... is it noraml...