Viewpoint-Conditioned Legible Motion Planning with Imitation and Reinforcement Learning AI and Manipulation