-
Notifications
You must be signed in to change notification settings - Fork 68
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
Array size mismatch when running predict_ros.py
#57
Comments
Update, we figured out the array-mismatch issue: iros20-6d-pose-tracking/predict_ros.py Lines 26 to 27 in d645214
The topics seem to be flipped around. (depth image topic for colour, and colour image topic for depth), putting them in the right places seem to fix it. However, one part of the original issue where the corp produces a completely empty image, 0 by 0 pixels, seems to remain?
|
I do notice that the |
Yes, the pose initialization is not included by se(3)-tracknet, which only tracks the pose afterwards. |
We run with slight modification to the logging to allow us see what is running under the hood. Like this one at start of
on_track
:This is happening on our own training data as mentioned in #55. Is this an issue with our data? Or is it an issue caused by Open3D, as we get an warning regarding the model not being read in due to unknown file format.
Or is it an issue with our Realsense camera? The model we are using is a D435. Here is an echo of from our
realsense/aligned_depth_to_color/camera_info
if it helpsThe encoding of the messages from the aligned-depth-to-color topic is 16UC1, same for the depth.
The text was updated successfully, but these errors were encountered: