Curious about dataset's efficacy in CSLAM research

[Mechazo11]

Hello CoPeD team.

Thank you very much for releasing this fantastic dataset and making it open source for research.

I am very much interested in using this dataset for a CSLAM project that I am currently working on. In this regard I have the following questions

1. Are the depth map generated by the Zoe depth in metric scale available?
2. Do the robot's have ground-truth poses and are they recorded in metric scale?
3. Are the class ids, 2D bounding boxes and semantic masks available per RGB frame for both UAV and UGV.

Thank you for your time.

With best regards,
@Mechazo11

[makeecat]

1. The depth map generated by the zoedepth should be with metric scale, but we are working on improving the depth result in the next six months.
2. The ground robots have pretty accurate GPS ground truth pose, and they are recorded in metric scale. We are also working on improving it, and we plan to release improved ground truth by the end of October.
3. These data are available already in the provided dataset.