Learning to grasp everyday objects using reinforcement-learning with automatic value cut-off | IEEE Conference Publication | IEEE Xplore