Generate a detailed image caption with highlighted entities
Segment objects in images and videos using text prompts
