Action-Driven Object Detection with Top-Down Visual Attentions
2016·Arxiv