SporeAgent - Reinforced Scene-level Plausibility for Object Pose Refinement

Dominik Bauer, Timothy Patten, Markus Vincze

Abstract:

Observational noise, inaccurate segmentation and ambiguity due to symmetry and occlusion lead to inaccurate
object pose estimates. While depth- and RGB-based pose
refinement approaches increase the accuracy of the resulting pose estimates, they are susceptible to ambiguity in the
observation as they consider visual alignment. We propose to leverage the fact that we often observe static, rigid
scenes. Thus, the objects therein need to be under physically plausible poses. We show that considering plausibility
reduces ambiguity and, in consequence, allows poses to be
more accurately predicted in cluttered environments. To this
end, we extend a recent RL-based registration approach towards iterative refinement of object poses. Experiments on
the LINEMOD and YCB-VIDEO datasets demonstrate the
state-of-the-art performance of our depth-based refinement
approach. See github.com/dornik/sporeagent