Face Recognition based Attendance Management System (FRAMS)

Face Recognition based Attendance Management System with a Flask web application and Power BI attendance dashboard.

Table of Contents

• Features
• Installation and Usage
• Technologies Used
• Methodology
• User Interface Demo

Features

• Face detection and recognition
• Attendance management
• Generates attendance reports in a csv file
• Secure admin login
• Interactive user interface
• Can detect multiple faces and mark attendance at a time
• Works in bright and low light conditions
• Attendance dashboards using Power BI

Installation and Usage

1. Install the required dependencies:

   pip install -r requirements.txt
   

2. Replace the training images with your own set of images in the folder Training images.
3. Open the app.py file and change the file paths as per your system.
4. Run the app.py file.

Technologies Used

• Programming Languages: Python
• Libraries: OpenCV, dlib, face-recognition
• Database: SQLite
• Web Application: Flask, HTML, CSS, JavaScript
• Data Visualization: Power BI

Methodology

• Environment Setup: Created a conda environment and installed necessary dependencies including OpenCV, dlib, face-recognition, and Flask.
• Face Detection: Converted images to black and white, then used HOG to detect faces by comparing image gradients.
• Face Embedding: Used 128-dimensional vectors and the triplet loss function for distinguishing between faces.
• Face Recognition: Utilized Euclidean distance with a threshold of 0.5 to compare the generated face encodings with the actual encodings of the training images to recognize the faces.
• Database Connection: Stored attendance data in a SQLite database and exported it to CSV for Power BI integration.
• Web Application: Developed a Flask-based web app for real-time attendance capturing and management.
• Power BI Dashboard: Connected the attendance data to Power BI to create dashboards. Embedded Power BI reports into the web app for real-time insights.