Segment Anything
Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alexander C. Berg, Wan‐Yen Lo, Piotr Dollár, Ross Girshick
2023 · 10,762 citations
Abstract
We introduce the Segment Anything (SA) project: a new task, model, and dataset for image segmentation. Using our efficient model in a data collection loop, we built the largest segmentation dataset to date (by far), with over 1 billion masks on 11M 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. We evaluate its capabilities on numerous tasks and find that its zero-shot performance is impressive – often competitive with or even superior to prior fully supervised results. We are releasing the Segment Anything Model (SAM) and corresponding dataset (SA-1B) of 1B masks and 11M images at segment-anything.com to foster research into foundation models for computer vision. We recommend reading the full paper at: arxiv.org/abs/2304.02643.
Cite this paper
Kirillov, A., Mintun, E., Ravi, N., Mao, H., Rolland, C., Gustafson, L., Xiao, T., Whitehead, S., Berg, A. C., Lo, W., Dollár, P., & Girshick, R. (2023). Segment anything. 3992–4003. https://doi.org/10.1109/iccv51070.2023.00371
Read it with every claim anchored
Add this paper to a project, ask questions of it, and get answers that point to the exact passage.
Start freeRelated papers
- You Only Look Once: Unified, Real-Time Object Detection2016
- A model of saliency-based visual attention for rapid scene analysis1998
- Frequency-tuned salient region detection2009
- Distributed and Overlapping Representations of Faces and Objects in Ventral Temporal Cortex2001
- Representational similarity analysis – connecting the branches of systems neuroscience2008
Metadata from OpenAlex (CC0). Citations are generated from the published record.