About This Architecture

SafeEntry is a Progressive Web App (PWA) attendance system combining face recognition, role-based authentication, and offline-capable deployment via AWS and Google Colab. The architecture layers a Flask REST API backend with InsightFace and OpenCV face detection, a PWA frontend supporting admin, teacher, and student portals, and JSON-based data storage for profiles and attendance records. Authentication gates all endpoints with session validation, while the face recognition engine preprocesses webcam images, detects faces using primary and fallback models, generates embeddings, and scores similarity against stored profiles. Developers can fork this diagram to customize slot management, add database backends, or integrate with institutional identity systems. The modular design separates concerns across auth, slot management, dashboard, and self-service APIs, making it ideal for educational institutions seeking biometric attendance without vendor lock-in.

People also ask

How do you build a face recognition attendance system with offline support and role-based access for teachers and students?

SafeEntry combines a Flask REST API backend with InsightFace and OpenCV for face detection and embedding, a PWA frontend with offline caching via service workers, and role-based authentication gates (admin, teacher, student). Webcam images are preprocessed using CLAHE enhancement, faces are detected with InsightFace Buffalo_L (primary) or Haar Cascade (fallback), embeddings are generated, and cosi