I need a complete, production-ready face-recognition attendance solution that works with three networked cameras: • Camera 1 at a registration kiosk for high-quality facial enrolment • Camera 2 monitoring the entry gate • Camera 3 watching the exit gate As people move through the doors the system must pick them up in real time, log their movement, and drive a 3-M-style security gate. Should face recognition fail, the gate must still open via either a scannable QRCode or a fingerprint reader—both backup methods must be fully supported. Inside the web dashboard I want to watch our building live: a rolling “Users Inside” counter that updates instantly. Historical analytics are equally important, so the dashboard also needs: • Daily entry and exit logs searchable by name, date, or door • Monthly attendance summaries with export to CSV/PDF The face engine should learn continuously to improve accuracy; please state in your proposal which frameworks (e.g., OpenCV, TensorFlow, FaceNet) you intend to use and how you will harden them against spoofing. The final package must include source code, build scripts, hardware setup instructions, and a short user manual so my IT team can extend it later. Acceptance testing is complete when the three cameras, QRCode scanner, and fingerprint module reliably trigger the gate and every analytic view displays the correct counts.