Face Recognition Attendance Management System (FRAS)

FRAS is the final-year graduation project I developed last year for university. It was built to demonstrate my practical software development skills in a real academic project context, including requirements analysis, system design, implementation, and testing. The project focuses on solving a real problem with a user-friendly interface, structured workflows, and reliable data handling, and it reflects my learning progress during the final year of study.

An automated, real-time biometric attendance management system built with Python, Django, MTCNN, and TensorFlow/Keras. FRAS eliminates manual roll-call processes by leveraging computer vision pipelines to detect faces from live video streams, extract facial feature embeddings, match identities against registered profiles, and log timestamped records automatically into a central web database.

---

Key Features

• Automated Detection & Alignment: Uses Multi-task Cascaded Convolutional Networks (MTCNN) for precise face detection and alignment under varying lighting conditions.
• Deep Learning Recognition: Extracts high-dimensional facial embeddings using TensorFlow and Keras pre-trained models for identity verification.
• Real-Time Stream Processing: Processes live camera frames in real time with low latency to detect and label subjects.
• Django Administrative Portal: A web dashboard to register new users, manage profiles, monitor live recognition sessions, and analyze attendance history.
• Automated Logging & Reporting: Prevents duplicate logs per session and saves timestamped attendance entries directly to a database.

---

System Architecture & Pipeline

[ Video Input Feed ]
         │
         ▼
[ MTCNN Face Detection ]       ───► Detects Bounding Boxes & Facial Landmarks
         │
         ▼
[ Deep Feature Extraction ]    ───► Generates Embeddings (TensorFlow / Keras)
         │
         ▼
[ Vector Matching Engine ]     ───► Compares Vectors with Stored Profile Data
         │
         ▼
[ Django Backend & Database ]  ───► Records Timestamped Attendance Entries

Tech Stack & Dependencies

• Programming Language: Python 3.8+
• Web Framework: Django
• Computer Vision & Processing: OpenCV, MTCNN
• Deep Learning Libraries: TensorFlow, Keras, NumPy
• Database: SQLite (Development) / PostgreSQL (Production)

Project Structure

├── manage.py
├── requirements.txt
├── README.md
├── fras_core/              # Django core settings and URL configurations
│   ├── settings.py
│   ├── urls.py
│   └── wsgi.py
├── attendance_app/         # Main app logic
│   ├── models.py           # Database models for user profiles and logs
│   ├── views.py            # Stream ingestion and administrative views
│   ├── utils.py            # Computer vision and model inference routines
│   ├── urls.py
│   └── templates/          # HTML templates for the web portal
└── dataset/                # Stored user profile vectors and images

Complete Installation & Setup Guide

Follow these steps sequentially to set up and run the system locally on your machine:

1. Clone the Repository

• git clone https://github.com/your-username/face-recognition-attendance.git
• cd face-recognition-attendance

2. Set Up a Virtual Environment

On Linux / macOS:

• python3 -m venv venv
• source venv/bin/activate

On Windows:

• python -m venv venv
• venv\Scripts\activate

3. Install Required Dependencies

Ensure you have pip updated, then run:

• pip install --upgrade pip
• pip install -r requirements.txt

4. Apply Database Migrations

Create the database tables for user profiles, attendance logs, and sessions:

• python manage.py makemigrations
• python manage.py migrate

5. Create an Administrative Superuser

Create an admin account to access the administrative web portal:

python manage.py createsuperuser

6. Run the Development Server

Start the Django server:

python manage.py runserver

Open your browser and go to http://127.0.0.1:8000/.

How to Use the System

1. Log In: Access the Django admin portal using your superuser credentials.

2. Register Users: Create user profiles (e.g., students or employees) and collect face sample images to build baseline feature embeddings.

3. Start Live Tracking: Launch the recognition view from the web portal to open the camera stream.

4. Automatic Logging: As registered individuals appear in front of the camera, the system detects their faces, verifies their identity, and creates an attendance record with a timestamp in the database.

5. View Reports: Access the dashboard to view daily logs, filter by date, or review missing entries.