Ask for advice on testing the method on our own data (without ground truth)

Hi, thank you very much for sharing this excellent work and the open-source code!

I am currently trying to test the method on my own data, and I would like to ask for your advice on whether and how this is supported.

Specifically, for each sample I only have:
- normal_map.png (estimated surface normal map)
- mask.png (valid region mask)
- K.txt (camera intrinsic matrix)

I do not have ground-truth normal maps or depth maps for evaluation.

I would greatly appreciate any guidance or pointers.
Thank you again for making this work publicly available!