The task of determining how well an image corresponds to a text description, often by computing a similarity score between them.
Quality of vision, audio, and image understanding (distinct from modality support)