1 paper
Zhekun Luo, Shalini Ghosh, Devin Guillory +3
Action in video usually involves the interaction of human with objects. Action labels are typically composed of various combinations of verbs and nouns, but we may not have trainin…