1 paper · 1 filter
Caihua Liu, Xu Li, Wenjing Xue +2
Existing video captioning methods merely provide shallow or simplistic representations of object behaviors, resulting in superficial and ambiguous descriptions. However, object beh…