Inverse Constraint Learning and Generalization by Transferable Reward Decomposition | IEEE Journals & Magazine | IEEE Xplore