Miguel Riem de Oliveira

Results 38 issues of Miguel Riem de Oliveira

Right now all joints affect all residuals. That could be improved.

enhancement

In #821 we've shown that using multiple patterns actually works in improving calibration. That was a single experiment. We should run a batch of experiments to make sure.

enhancement

When running rosrun atom_calibration generate_joint_space_tsne -json ./dataset.json I get this error ``` Traceback (most recent call last): File "/home/mike/workspaces/catkin_ws/src/calibration/atom/atom_calibration/scripts/utilities/generate_joint_space_tsne", line 76, in main() File "/home/mike/workspaces/catkin_ws/src/calibration/atom/atom_calibration/scripts/utilities/generate_joint_space_tsne", line 40, in main raise...

bug

In some places such as the transforms we are using rodrigues, some others Euler angles. We should make it consistent.

enhancement

https://github.com/lardemua/atom/blob/1603e3cc61650104f269a42b911c422787c930a4/atom_calibration/src/atom_calibration/calibration/objective_function.py#L390

enhancement

[ERROR] [1701803905.669482110]: PluginlibFactory: The plugin for class 'rviz_plugin_selected_points_publisher/SelectedPointsPublisher' failed to load. Error: Could not find library corresponding to plugin rviz_plugin_selected_points_publisher/SelectedPointsPublisher. Make sure the plugin description XML file has the correct...

bug

Texture is wrong: ![image](https://github.com/lardemua/atom/assets/11350998/2602cb3d-a504-4600-9fa8-b76d773a819c)

bug

After optimizing this system, it seem like we have good results: ![image](https://github.com/lardemua/atom/assets/11350998/5352c72e-6a47-4fec-9efc-df93977166f5) But the sensors do not appear to be well matched in rviz: ![image](https://github.com/lardemua/atom/assets/11350998/a9b31d02-38ed-47c2-b336-2dad11c6ee4f)

enhancement

fail if not transform_ground_truth exist in the dataset ...

enhancement

vision.py imports o3d which then gives an error concerning the numpy versions. Often we do not need the functions that use o3d, so let's split the vision.py into two files,...

enhancement