The documentation mentions the cuVGL benchmark metrics:
Is there a way to reproduce these benchmark results on my own setup?
I looked into the benchmark node here:
—but I couldn’t find specific instructions or configurations related to cuVGL benchmarking. Any pointers or guidance would be appreciated.
The documentation also mentions the topic /visual_localization/pose.
In my current setup, I only receive a pose with zero covariance values.
Is this expected behavior for the pose estimation, or is something misconfigured?
