論文サマリ

acl2020
tag: nlvr

Cross-Modality Relevance for Reasoning on Language and Vision
by: sobamchan
multi-modal transformer vqa nlvr