Segment Anything Model

The dataset used in this paper is the Meta Research's Segment Anything Model (SAM) dataset, which consists of images.

BibTex: