Vision-Language Grounding as Bidirectional Concept Correspondence

Part of

Vision-Language-Action And Multimodal Models