A machine learning approach that learns to match text descriptions with images by training them in a shared semantic space, enabling tasks like image search and visual understanding.
Quality of vision, audio, and image understanding (distinct from modality support)