kmvishn/realtime-weapon-detection — reverse-engineered prompt
Reverse engineered prompt
Build me a simple real time weapon detection app in Python that can watch a live camera feed, spot weapons, and immediately raise an alert. I want a desktop window for monitoring the video and a basic web interface too, so I can view detections and manage things from a browser. When something is detected, play an alarm sound and make the alert easy to customize.
Please use the included model files if they already work, and keep the whole project practical and easy to run locally. I also want the training and testing parts wired up so I can retrain on new images later and check how the model performs. If there is already a script for downloading images, connect that into the workflow instead of reinventing it.
Keep the code organized so the detection logic, the desktop app, and the web server are separate but work together. Add clear setup steps and make sure I can start the GUI and the web app without guessing. If you need to, look up current docs online.
Want more depth? Deep Reverse