A Bayesian Approach to Reinforcement Learning of Vision-Based Vehicular Control | IEEE Conference Publication | IEEE Xplore