difference in generated depth ground truth and kitti depth annotations

[sanweiliti]

Hi,
Depth ground truth for testing in this code are generated from some information and parameters from the kitti raw data, while some other papers used the kitti depth annotations provided by kitti depth benchmark for testing and reporting results. Will there be much difference in these two types of depth ground truths and test results?

Thanks!