Segment Anything

The Segment Anything Model (SAM) is an AI model created by Meta AI which can generate masks for different objects in images. It is trained on a dataset consisting of 11 million images and 1.1 billion masks, giving it good performance on a variety of segmentation tasks. This new model is accompanied by the SA-1B dataset, which has 400 times more data than other existing segmentation datasets and is intended to help further computer vision research and develop new applications.