Back to Newswire
Science

Meta AI: SAM 3 and DINOv3 cut DOE Genesis Mission 3D volume labeling from ~1 month to ~15 minutes

Meta AI announced that SAM 3 and DINOv3 are being used by the Berkeley Lab-led SYNAPS-I project to automate image segmentation for the Energy Department's Genesis Mission. The post said that by pairing DINOv3's global semantic context and fine-grained spatial localization with SAM 3's pixel-level boundary extraction, researchers are able to compress 3D volume labeling from a month of manual effort to about 15 minutes. The post included a link to learn more about their work.
Sources
AIatMeta
In this story
Meta US Government
Published by Tech & Business, a media brand covering technology and business. This story was sourced from AIatMeta and reviewed by the T&B editorial agent team.
Related Stories
Science

University of Waterloo team 3D-prints custom contact lenses in about 20 minutes

New Atlas
Science

WHO-commissioned reviews find no link between cell phones and brain cancer

The Conversation via Refractor
Science

Greenland meltwater adds major AMOC weakening but not irreversible shutdown, model study finds

Live Science
Science

ETH Zurich chip uses mechanical vibrations as quantum working memory, Science study shows

SciTechDaily