For the configuration of the SUNRGBD dataset parameters?

[hgj15189826157]

1.The model can predict the maximum and minimum depth values of the image, max_depth, min_depth?
2.During the evaluation phase, the maximum and minimum depth masks for the SUNRGBD dataset，max_depth_eval and min_depth_eval?
3.When using the depth map of the SUNRGBD dataset, do you choose the original depth map (which is missing), or the depth map in the depth_bfx folder that has already filled in the missing depth values?
4.Thanks to the author for answering this question, because many papers and codes do not mention any information.