Implementation Details for datasets

[pp00704831]

I saw that you used datasets GoPro, DVD, and NFS for training in your paper.
Does it means that I need download that three datasets mentioned above, and put all of them in following way?
train:
files_a: ./datasets/my_dataset/train//blur/*.png
files_b: ./datasets/my_dataset/train//sharp/.png
...
...
val:
files_a: ./datasets/my_dataset/test/**/blur/.png
files_b: ./datasets/my_dataset/test/**/sharp/*.png
...
...
After that, I can just use command : python train.py, then datasets setting will be the same as paper?
Thank You!