Improved reward estimation for efficient robot navigation using inverse reinforcement learning | IEEE Conference Publication | IEEE Xplore