Visual Affordance Prediction for Guiding Robot Exploration