Self-Attention Based Visual-Tactile Fusion Learning for Predicting Grasp Outcomes | IEEE Journals & Magazine | IEEE Xplore