An Azure Vision feature that writes a single readable sentence describing an image; dense captions instead describe as many as 10 regions.
Read more: Microsoft Learn
In the Ultra Transcenders books
Each book explains Image captioning in context, with comparison tables and the common traps.
Terms in this definition
- Description (Image Analysis 3.2)
Visual feature in Image Analysis 3.2 that produces natural-language caption sentences sorted by confidence, read from Description.Captions[0].Text. Version 4.0 swaps it for Caption.