Multimodal QA

Visual Grounding

The act of anchoring linguistic concepts (words, phrases) to specific entities or regions in an image or video, creating a tangible semantic link.

← Wstecz