Ourmodel is trained in an EM-type manner alternating between increasing the 3D pose invariance ofthefeature extractor andannotating unlabelled data through neural viewsynthesis andmatching.

Robust reinforcement learning (RL) is to find a policy that optimizes the worstcase performance over an uncertainty set of MDPs. In this paper, we focus on model-freerobust RL, where the uncertainty set is defined to be centering at a misspecified MDP that generates a single sample trajectory sequentially, and is assumed to beunknown.

However, when merging a parameter that is influential for one model but redundant (i.e.