[ICLR 2022] RelViT: Concept-guided Vision Transformer for Visual Relational Reasoning
Primary LanguagePythonOtherNOASSERTION