An attention-driven hierarchical multi-scale representation for visual recognition

Open in new window