A.U.R.A — Attendance & User Recognition Application

A.U.R.A (Attendance and User Recognition Application) is a smart attendance management system that combines RFID authentication and facial recognition to automate and secure attendance tracking in educational institutions.

---

Key Features

• RFID-based student authentication
• Facial recognition for proxy prevention
• Automated attendance logging
• Faculty dashboard for monitoring
• Secure database storage
• Real-time verification

---

Project Structure

A.U.R.A/
│
├── app/ # Application entry (frontend)
├── assets/ # Images, icons, and static files
├── components/ # Reusable UI components
├── hooks/ # Custom React hooks
├── api.py # Backend API (Python)
│
├── app.json # App configuration
├── tsconfig.json # TypeScript configuration
├── package.json # Node dependencies
├── package-lock.json # Dependency lock file
├── loading-circle.json # Loader animation
├── .prettierrc # Code formatting rules
├── .gitignore # Ignored files
└── README.md # Project documentation

---

Tech Stack

Frontend

• TypeScript
• React Native / Expo
• Custom UI Components

Backend

• Python
• FastAPI (API Layer)

Database

• MongoDB

Hardware (Prototype)

• Arduino Uno
• RFID Reader
• Raspberry Pi
• Camera Module / Mobile Camera

Machine Learning

• MTCNN (Face Detection)
• FaceNet (Embeddings)
• SVM Classifier

---

System Architecture & Pipelines

The A.U.R.A system is built around two core pipelines that work together to ensure accurate and secure attendance tracking:

• Model Training Pipeline — Responsible for preparing and training the face recognition model.
• Application & Attendance Pipeline — Handles real-time attendance collection, verification, and reporting.

---

Model Training Pipeline

This pipeline is used to train the facial recognition system using classroom images.

Workflow

1. Classroom images are collected during sessions
2. Faces are detected using MTCNN
3. Individual faces are extracted from group images
4. Data augmentation is applied (rotation, cropping, flipping, etc.)
5. Face embeddings are generated using FaceNet
6. Embeddings are classified using SVC (Support Vector Classifier)
7. The trained model is saved for deployment

Architecture Diagram

This pipeline improves recognition accuracy by learning robust facial features under different lighting and viewing conditions.

---

Application & Attendance Pipeline

This pipeline manages live attendance collection and proxy detection in classrooms.

Workflow

1. Faculty logs in through the dashboard
2. User authentication is performed
3. RFID data is collected from entry devices
4. Classroom images are captured using cameras
5. Images are sent to the GPU-enabled model server
6. Face recognition is performed
7. Proxy verification is applied
8. Attendance is finalized and stored in the database
9. Reports are generated and exported (Excel format)

Architecture Diagram

This pipeline integrates IoT devices, AI inference, and backend services to provide real-time and reliable attendance monitoring.

---

Conclusion

A.U.R.A provides an efficient and reliable solution for automated attendance management by combining RFID technology with facial recognition. The system reduces manual effort, prevents proxy attendance, and ensures accurate record keeping. With its scalable architecture and user-friendly interface, A.U.R.A can be effectively used in educational institutions to improve attendance monitoring and management.