1 paper
Hsin-Chen Pai, Jyun-Kai Wang, Yi-Cheng Peng +1
Referring multi-object tracking (RMOT) takes a video and a language expression as input and tracks all referred objects. Many tracking requirements involve how an object moves rath…