Meta Launches Pic Recognition AI Tool
After its Metaverse goals failed, Meta now relies on AI to advance its development. Meta launched the Segment Anything Model on Wednesday. With a few clicks, Segment Anything enables people to find objects in images. Segment Anything, which is still in beta, can take a snapshot and identify each pixel in an image, allowing one or more things to be separated from the rest.
“Segmentation—identifying which image pixels belong to an object—is a core task in computer vision and is used in a wide range of applications, from analysing scientific imagery to editing photos,” Meta said in announcing the new model. To create reliable segmentation models for specific jobs, technical professionals need AI training infrastructure and large amounts of precisely annotated in-domain data, according to Meta.
The Segment Anything AI system, according to Meta, was trained on 11 million photos. Meta is making Segment Anything available to researchers through the Segment Anything Github under a liberal open licence, Apache 2.0.
(With inputs from Shikha Singh)
You need to login in order to Like
