Closed-loop object recognition using reinforcement learning | IEEE Journals & Magazine | IEEE Xplore