Explore the zero-shot detection and segmentation capabilities of YOLOR, all in a user-friendly notebook setting. We’ll walk through setting up a GPU environment, installing key libraries, and running predictions on images and video clips. You’ll see how to apply text-based labels or bounding box prompts to handle tasks like identifying dogs, suitcases, or even croissants. Along the way, we’ll discuss environment configuration, label annotation, and methods to streamline detection for real-time performance. Ideal for those who want to dive into advanced object recognition without heavy retraining.

⭐ YOLOE: github.com/THU-MIG/yoloe
⭐ supervision: github.com/roboflow/supervision

Stay updated with the projects I'm working on at github.com/roboflow and github.com/SkalskiP! ⭐