Currently perfect depth from depth camera and lidar in simulator can overly simplify tasks when trying to make system applicable for sim-to-real transfer experiments.
Either implement own noise model or look out for additional noise properties being added to Omniverse in future releases (alluded to here) and see how they might be incorporated into benchbot framework task
Currently perfect depth from depth camera and lidar in simulator can overly simplify tasks when trying to make system applicable for sim-to-real transfer experiments.
Either implement own noise model or look out for additional noise properties being added to Omniverse in future releases (alluded to here) and see how they might be incorporated into benchbot framework task