Two questions about Depthlab

[Yang-SyZng]

1. In the paper "Table 1. Quantitative comparison of various methods on different datasets." AbsRel and δ1 are used to evaluate the performance of various depth models. As we all know, Depth Anything v2 and other methods only need one image to estimate the depth, so the estimated depth is scale-ambiguous, while the depth estimated by DepthLab contains the absolute scale. How is the quantitative comparison in this table done? In other words, how to unify the scale estimated by other models with DepthLab?
2. Another question is when generating the depth npy file, what unit is this depth in? Millimeters? When I use this file, the depth value needs to be divided by 2000 so that the point cloud can coincide with the actual collected point cloud.

[Johanan528]

Hello.:-)

First, in traditional monocular depth estimation, the process for evaluating metrics is as follows: initially, relative depth is estimated, and then a least squares transformation is applied to convert it to absolute depth based on the ground truth (gt) values provided in the benchmark. This is because relative depth alone cannot be directly used for evaluation. For more details, you can refer to this link: https://github.com/prs-eth/Marigold/blob/main/eval.py.

It is worth noting that our method does not require the least squares transformation. Instead, we only need to remember the normalization parameters used during input, such as the maximum and minimum values, to easily convert the depth back to absolute depth.