Segment Anything

The Segment Anything (SA) project introduces a new task, model, and dataset for image segmentation. Using their efficient model in a data collection loop, the authors built the largest segmentation dataset to date by far, with over 1 billion masks on 11 million licensed and privacy-respecting images. The model is designed and trained to be promptable, so it can transfer zero-shot to new image distributions and tasks.

Evaluated on numerous tasks, the model's zero-shot performance is impressive, often competitive with or even superior to prior fully supervised results. The Segment Anything Model (SAM) and the corresponding SA-1B dataset of 1 billion masks and 11 million images are released at segment-anything.com to foster research into foundation models for computer vision.