Nawaf-Rayhan585/YOLO_Projects ? reverse-engineered prompt

Build me a Python computer vision project that lets me run a few different YOLO based demos from one place.

I want scripts for things like detecting people, cars, fire, rats, hands, and human presence, plus a people counting mode that can track in and out movement in a zone, a crowd heatmap view, a suspicious behavior detector using pose and YOLO, and a reels distraction detector for someone looking at a phone. Each one should work on a webcam, video file, or image when it makes sense, and show the results live with boxes, labels, confidence scores, and alerts on screen.

Keep it simple to run, with a clear main entry point and easy to understand code. If you need current YOLO or OpenCV docs while wiring things up, feel free to look them up. Make it feel like a real usable demo set, not just random notebook experiments, and include basic setup instructions so I can get it running quickly.