SAM 3: Prompt once, segment faster with new annotation workflows in Encord

With the release of Meta's SAM 3 model, auto-segmentation and object detection is faster than ever - but how can AI teams leverage this for faster data annotation?

In this live session, Frederik Hvilshøj and Alex Bonnet will discuss the significance of the SAM 3 model, how AI teams can use auto-segmentation to cut annotation time, and what these workflows look like in Encord.

You’ll learn how to:

• Use text prompts to auto-segment multiple objects in images and video
• Track objects in videos forwards and backwards through time
• Accelerate your end-to-end annotation workflows with SAM 3
• Build SAM 3 workflows in the Encord platform

Who it’s for:

Machine learning engineers, annotation teams, and data ops professionals who want to streamline their video and image annotation workflows with SAM 3.