Asymmetric Self-Play for Learning Robust Human-Robot Interaction on Crowd Navigation Tasks | IEEE Conference Publication | IEEE Xplore