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

Centerpoint cannot detect any objects for livox mid 360 #10317

Open
3 tasks done
leoxiang66 opened this issue Mar 21, 2025 · 2 comments
Open
3 tasks done

Centerpoint cannot detect any objects for livox mid 360 #10317

leoxiang66 opened this issue Mar 21, 2025 · 2 comments
Labels
component:perception Advanced sensor data processing and environment understanding. (auto-assigned)

Comments

@leoxiang66
Copy link

Checklist

  • I've read the contribution guidelines.
  • I've searched other issues and no duplicate issues were found.
  • I'm convinced that this is not my fault but a bug.

Description

Hi guys, I tried both centerpoint and centerpoint_tiny models but both cannot detect any objects.

I see on this issue that the training data matters (e.g. different beams or scan mode).

I would ask to see if anyone has relevant experience on centerpoint for mid 360?

Many thanks!

Expected behavior

detecting objects

Actual behavior

detect nothing

Steps to reproduce

normal steps

Versions

ubuntu2204
ros2 humble

Possible causes

No response

Additional context

No response

@idorobotics idorobotics added the component:perception Advanced sensor data processing and environment understanding. (auto-assigned) label Mar 21, 2025
@knzo25
Copy link
Contributor

knzo25 commented Mar 24, 2025

@leoxiang66
If centerpoint can not detect "any" objects it is probably not due to the lidar itself but rather data format, environment it was tested and so on. If you can provide a rosbag, we can check it later

@leoxiang66
Copy link
Author

@knzo25
Hi Knzo25, thanks for reply!

You can find the rosbag here: https://drive.google.com/file/d/1y3UIEVVIhrxvtQM4CJ7HMVeFMlaBl3od/view?usp=sharing

Btw, I have also tested inferencing on the original training dataset (nuscenes), and centerpoint can detect objects. So I doubt it may be the different distribution of these two datasets.

May you have a test on this rosbag too? I would appreciate it!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
component:perception Advanced sensor data processing and environment understanding. (auto-assigned)
Projects
Status: To Triage
Development

No branches or pull requests

3 participants