Fetching the paper…

RLIP: Relational Language-Image Pre-training for Human-Object Interaction Detection · Around