Facebook put more advancement in improving the photo experience for its users. Expanding the number of concepts that AAT (Automatic Alternative Text) can detect and identify in a photo by more than 10x; more detailed descriptions with the ability to identify activities, landmarks, types of animals, and so forth and the capability to include information about the positional location and relative size of elements in a photo help blind or visually impaired users to better understand what’s in the photos posted by their family, friends and even themselves.
Pillars of AI research and area of focus for Facebook such as image recognition, object detection using Detectron2 and transfer learning designed the new AAT to provide a succinct description for all photos by default but offer an easy way to get more detailed descriptions about photos of specific interest and to benefit people in the real world.
