1 paper
Runhao Zeng, Jiaqi Mao, Minghao Lai +5
Video grounding (VG) task focuses on locating specific moments in a video based on a query, usually in text form. However, traditional VG struggles with some scenarios like streami…