Virtual-to-real deep reinforcement learning: Continuous control of mobile robots for mapless navigation | IEEE Conference Publication | IEEE Xplore