Nawaf-Rayhan585/Reels-Detection-Model-YOLOv26 — reverse-engineered prompt

Build me a simple Python project for real time distraction detection using the latest YOLOv26 model. I want to be able to run one script and have it open my webcam or a video file, detect distracting objects or situations on screen, and draw clear boxes and labels with a confidence score. It should feel like a clean demo that is easy to test, not a huge complicated app.

Please make it practical and beginner friendly. Add the model loading, video capture, live inference, and a simple way to quit the window. If it makes sense, include a small option to save the processed output video too. I also want a short README that explains what the project does, how to install it, how to run it, and where I would swap in my own trained weights if needed.

Keep the code organized and readable in one main script if that is enough. Look up current YOLO docs online if you need to.