Ultralytics YOLO11 | Training, Inference, Benchmarking, and Deployment Explained! 🚀 - How to Use Meta Segment Anything 3 with Ultralytics | Text-Prompt Segmentation on Images & Videos 🚀

Meta recently released Segment Anything 3 (SAM3). In this tutorial, we walk through how to use SAM3 with the Ultralytics framework to perform segmentation on images and videos using text prompts. You’ll learn how to run inference with SAM3 using single and multiple text prompts on images, followed by video-based segmentation driven by text prompts. The tutorial focuses on understanding how text-prompt segmentation works in practice and how SAM3 can be applied to tasks such as specific object segmentation, interactive segmentation, and other computer vision workflows. Chapters: 00:00 - Introduction to Meta Segment Anything 3 (SAM3) 01:03 - SAM3 documentation walkthrough 03:31 - Installing the Ultralytics Python package 04:44 - Segmentation with SAM3 in an image using a text prompt 08:49 - Segmentation with SAM3 using multiple text prompts 10:12 - Segmentation with SAM3 in a video using text prompts 10:50 - SAM3 video inference output demo 11:35 - conclusion and key takeaways 🔗 SAM3 documentation ➡️ https://docs.ultralytics.com/guides/streamlit-live-inference/ 🔗 SAM3 code (used in tutorial) ➡️ https://drive.google.com/file/d/1mjRy6VmHke95zdkaCmR_RWvPP4Hnba7q/view?usp=sharing Ultralytics Resources: 🏢 About Us: https://ultralytics.com/about 💼 Join Our Team: https://ultralytics.com/work 📞 Contact Us: https://ultralytics.com/contact 💬 Discord Community: https://discord.com/invite/ultralytics 📄 Ultralytics License: https://ultralytics.com/license #sam3 #segmentanything #ultralytics #computervision #segmentation #visionai #promptengineering