Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Curious about dataset's efficacy in CSLAM research #1

Open
Mechazo11 opened this issue Aug 15, 2024 · 2 comments
Open

Curious about dataset's efficacy in CSLAM research #1

Mechazo11 opened this issue Aug 15, 2024 · 2 comments

Comments

@Mechazo11
Copy link

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
Copy link
Collaborator

  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.

@Mechazo11
Copy link
Author

Hi @makeecat

I am writing today to follow up our initial conversation last August. In the arplaboratory/CoPeD dataset I recently saw the depth presented as videos. A sample from NYUARPL_ptah_depth_n.mp4 is shown below

Image

Is this an estimated depth map from ZoeDepthV2?

Best regards,
@Mechazo11

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants