Attention in Computer Vision, Part 3: GE