1 paper
Yicheng Qian, Weixin Luo, Dongze Lian +3
In this paper, we propose a novel sequence verification task that aims to distinguish positive video pairs performing the same action sequence from negative ones with step-level tr…