Meta presents an AI model for image segmentation

Meta introduces Segment Anything, a new model for image segmentation. A way to “crop” any object in a single click.

Meta has launched a new artificial intelligence (AI) model that has the ability to identify specific objects in an image, as well as a large volume of carefully annotated domain data. The Segment Anything Model, or SAM for short, can recognise elements in images and videos like never before.

To further democratise segmentation, Meta is making the massive SA-1B dataset available for research, and the Segment Anything model is available under an open licence (Apache 2.0). In addition, the company has developed a window for you to test what this new image and video segmentation model is capable of.

Meta’s augmented reality glasses will arrive in 2027.