[datasets] Availability of the datasets

[MarcSzafraniec]

Hi, thanks for this excellent work!

We would be interested in testing the robustness of our depth models on your ensemble of datasets. However it seems that some evaluation datasets are missing, in particular NYUDepth2-C and KITTI-S.

Do you plan to make these datasets accessible ?

The ground truth depths for the test set of the challenge would also be interesting, if you can release them.