Attendance Management System

Using Face Recognition

A Project Report

submitted in partial fulfillment of the requirements

of

AICTE Internship on AI: Transformative Learning

with
TechSaksham – A joint CSR initiative of Microsoft & SAP

by

Gopika M , gopikamuru47@gmail.com

Under the Guidance of

Aditya Prashant Ardak

Master Trainer , Edunet Foundation
 ACKNOWLEDGEMENT

I would like to express my sincere gratitude to everyone who supported me throughout the
course of this thesis work. Firstly, I am deeply thankful to my supervisor, Aditya Prashant
Ardak Sir, for his exceptional guidance and mentorship. His deep knowledge of chatbot
development and natural language processing has been instrumental in shaping my
understanding of the subject.

I would also like to thank Abdul Aziz Md for his valuable assistance with GitHub and
version control. His guidance was crucial in helping me manage the project effectively and
ensured smooth collaboration among the team. His expertise in organizing and maintaining
the project’s structure was invaluable.

Additionally, I am grateful to Pavan Sir for arranging insightful sessions His support has
been an integral part of the successful completion of this project.
 ABSTRACT

The Attendance Management System using Face Recognition is an advanced and automated
approach designed to streamline the process of tracking and managing student or employee
attendance. Traditional manual attendance methods are often time-consuming, prone to errors, and
vulnerable to proxy attendance. This system leverages state-of-the-art facial recognition technology
to identify individuals and record their presence automatically. The system works by capturing a
person's facial features and matching them with pre-stored images or a database of enrolled faces.
Upon successful recognition, the system logs the attendance without requiring any physical
intervention or manual entry. The proposed solution offers a high level of accuracy, speed, and
security while reducing administrative overhead. It also ensures that the attendance records are
tamper-proof and easily accessible for review or analysis. The system can be integrated into
classrooms, offices, or any organization that requires efficient attendance management. By using
deep learning techniques and image processing algorithms, this system aims to provide an efficient,
reliable, and user-friendly solution for attendance management.
 TABLE OF CONTENT

Abstract........................................................................................................................I

Chapter 1. Introduction..........................................................................................1
1.1 Problem Statement.............................................................................................1

1.2 Motivation..........................................................................................................1
1.3 Objectives..........................................................................................................1
1.4. Scope of the Project...........................................................................................1
Chapter 2. Literature Survey................................................................................2
Chapter 3. Proposed Methodology........................................................................3
Chapter 4. Implementation and Results...............................................................5
Chapter 5. Discussion and Conclusion................................................................10
References..................................................................................................................11
LIST OF FIGURES

Page
Figure No. Figure Caption
No.
Figure 1 User Registration 6
Figure 2 Face Detection 7

Figure 3 Attendance Marking 8

Figure 4 Database Management 9
 CHAPTER 1
Introduction

1.1 Problem Statement:

Traditional attendance management systems, such as manual registers or card-
based methods, are inefficient, time-consuming, and prone to human error. These
methods often result in inaccurate attendance records, delayed processing, and
opportunities for manipulation, such as proxy attendance. Additionally,
maintaining and managing physical records for large groups of students or
employees can be cumbersome and prone to data loss. There is a need for a more
efficient, accurate, and secure system that automates attendance tracking while
minimizing administrative overhead. The challenge is to develop a system that not
only simplifies the attendance process but also provides reliable verification of
individual identities to prevent fraud, reduces human involvement, and ensures
real-time, tamper-proof records. This project aims to address these issues by
utilizing face recognition technology, offering a solution that is fast, accurate, and
scalable for educational institutions or corporate environments.

1.2 Motivation:
The motivation behind developing an Attendance Management System using Face
Recognition stems from the growing need for efficient, accurate, and secure
methods of tracking attendance in educational institutions and workplaces.
Traditional methods of attendance management, such as manual entry or card-based
systems, are increasingly inadequate in addressing modern challenges. They are not
only time- consuming but also vulnerable to human errors, manipulation, and fraud,
such as proxy attendance. In large organizations or classrooms, these issues can
lead to administrative inefficiencies, lost productivity, and unreliable data.
By implementing such a system, institutions can also enhance security, provide
better user experience, and foster a more modern and tech-savvy environment. This
innovation aligns with the increasing reliance on automation and artificial
intelligence in education and business sectors.

1.3 Objective:

The objective is to develop an automated attendance management system using

face recognition technology to improve accuracy, prevent proxy attendance, and
streamline the attendance process. The system aims to reduce administrative
overhead, ensure real-time, tamper-proof records, and enhance security, providing a
pg. 1
 reliable and efficient solution for both educational and organizational environments.

1.4 Scope of the Project:

The scope of the Attendance Management System using Face Recognition

includes the implementation of an automated, real-time attendance logging system
based on facial recognition technology. It ensures secure, tamper-proof data storage
and offers a user-friendly interface for both administrators and users. The system will
be scalable for various group sizes, suitable for educational institutions and
workplaces. It also includes data analytics for attendance trends and can integrate with
existing platforms or databases. The solution ensures accuracy, efficiency, and
security in attendance management.

CHAPTER 2
Literature Survey

2.1 Review of Relevant Literature or Previous Work: Previous research indicates that
face recognition technology offers a highly efficient, accurate, and secure solution for
automated attendance management. However, it also highlights the challenges related
to system robustness, data security, and integration with existing infrastructures.
Building on these advancements, the proposed attendance management system seeks
to further enhance the reliability, scalability, and user experience of face recognition-
based solutions, addressing both technical and practical issues identified in earlier
works.
2.2 Current models like Symptom Checker and Existing Models, Techniques, or
Methodologies: The combination of advanced face recognition models, real-time
detection techniques, security measures, and cloud-based solutions offers a robust
foundation for developing an efficient and secure automated attendance
management system. These models and methodologies have been successfully used

pg. 2
in various

pg. 3
applications, and adapting them to the context of attendance management will ensure
accuracy, scalability, and reliability in the proposed system.
2.3 Gaps or Limitations in Existing Solutions: While face recognition systems for
attendance management offer a promising solution, existing models still face several
limitations, including issues related to accuracy, privacy, scalability, and integration
with other systems. Addressing these gaps will be critical in developing a robust,
secure, and efficient attendance management system that meets the needs of diverse
environments. Future advancements should focus on improving algorithm accuracy
under varying conditions, ensuring stronger data security measures, and optimizing
systems for scalability and user inclusivity.

CHAPTER 3
Proposed Methodology

3.1 System Design

1. User Interface (UI):

o Web/Mobile Application: Staff or admins can log in to view attendance
reports.

pg. 4
 o Live Feed UI: Shows the live camera feed where faces are captured.
o Report UI: Generates attendance reports and analytics.
o Tools: HTML, CSS, JavaScript (React or Angular).
2. Face Detection and Recognition Module:
o Face Detection: Captures faces from a live camera feed using Haar Cascade
or MTCNN.
o Face Recognition: Uses a model like FaceNet or Dlib for feature
extraction and comparing facial embeddings.
o Tools: OpenCV, TensorFlow, PyTorch, or Mediapipe.
3. Database Management:
o Employee/Student Database: Stores employee or student data, including
facial embeddings for comparison.
o Attendance Logs: Stores daily attendance logs with timestamps.
o Tools: MySQL, PostgreSQL, MongoDB.
4. Backend (Application Server):
o API Services: Expose REST APIs for recording attendance and managing
user data.
o Business Logic: Processes data, matches embeddings, and updates
attendance.
o Tools: Flask, Django (Python), Node.js (JavaScript).
5. Attendance Processing:
o Recognized faces are logged into the database with timestamps.
o Maintains records for absent and late arrivals.
6. Notification System (Optional):
o Sends alerts or attendance summaries via email or SMS.
o Tools: Twilio API or Email services.
7. Hardware Integration:
o Camera: Captures images or videos for face detection.
o Edge Devices: For real-time processing (like a Raspberry Pi or edge AI
devices).

3.1.1 Hardware Requirements:

CPU: A modern multi-core processor (e.g., Intel i5) for efficient task handling.

pg. 5
 RAM: Minimum of 8 GB RAM for running chatbots and managing session history
in real-time.
Internet: Stable internet connection for accessing Groq API and external resources.

3.1.2 Software
Requirements:
Operating System
 Windows / Linux / macOS

2. Programming Languages
 Python: For face recognition and backend.
 SQL: For the database.

3. Tools and Libraries

 Face Recognition:
o OpenCV (for face detection)
o Keras
o Pytorch
o tensorflow
 Database:
o MySQL (to store attendance records)

4. Development Tools
 IDE: Visual Studio Code
 Version Control: Git and GitHub

pg. 6
 CHAPTER 4
 Implementation and Result

4.1 Snap Shots of Result:

FIGURE 1 :

 This is the screen where new employees or students register their details and
face data.
 The system captures the user’s name, ID, and face image.
 The captured face is pre-processed and stored as face embeddings in the database
for future recognition.

pg. 7
FIGURE 2 :

 This step shows the system capturing a live feed from the camera.
 Faces in the live feed are detected using OpenCV
 The detected face is compared with stored face embeddings in the database using a
recognition model like FaceNet
 If a match is found, the system proceeds to mark attendance

pg. 8
FIGURE 3 :

 Once the face is recognized, attendance is automatically marked in the system.

 The system records details like Name, ID, Date, and Time of recognition.
 A confirmation message, like "Attendance Marked Successfully", is displayed.

pg. 9
FIGURE 4 :

 This screen shows the Attendance Report generated for a specific day, week, or
month.
 Reports can be viewed, filtered, and exported.
 Each row in the report shows the Name, ID, Date, Time, and Attendance Status
(Present/Absent)..

pg.
10
4.2GitHub Link for Code:

https://github.com/GOPIKAMURUGAN/Attendance-Management-
System.git
CHAPTER 5
Discussion and Conclusion

5.1 Future Work:

1. Integration of Advanced AI Models

 Improvement of Accuracy: Use more advanced deep learning models (e.g.,

Vision Transformers, improved CNNs) to enhance recognition accuracy,
especially in challenging conditions like low light or partial occlusions.
 Real-Time Processing: Optimize models for real-time performance,
reducing latency in attendance marking.
 Emotion and Expression Handling: Include capabilities to handle varied
facial expressions that could otherwise hinder recognition.

2. Hardware and Sensor Enhancements

pg. 10
  3D Face Recognition: Use 3D cameras and LiDAR technology for more
robust recognition that is less affected by lighting and pose variations.
 Edge Computing: Deploy AI models on edge devices like IoT-enabled cameras
for decentralized and efficient recognition.
 Wearable Devices: Develop mobile or wearable attendance systems
that integrate face recognition.

3. Multi-Modal Biometrics

 Fusion with Other Biometrics: Combine face recognition with

fingerprints, voice recognition, or iris scanning for increased accuracy and
reliability.
 Behavioral Biometrics: Incorporate behavioral patterns like gait
recognition to complement face data.

4. Cloud-Based and Distributed Systems

 Scalable Cloud Infrastructure: Design cloud-based solutions for scalability to

handle large institutions and enterprises.
 Hybrid System: Combine edge and cloud processing to balance speed
and computational power.
 Blockchain Integration: Use blockchain to ensure secure, tamper-proof
attendance records.

5. Anti-Spoofing and Security Improvements

 Liveness Detection: Enhance systems to detect spoofing attempts using

photos, videos, or masks through liveness detection techniques.
 Privacy-Preserving Face Recognition: Implement federated learning
or homomorphic encryption to preserve privacy while training models.
 Secure Data Transmission: Use encrypted communication protocols to
prevent unauthorized access to attendance data.

5.2 Conclusion:

The Face Recognition-Based Attendance Management System offers an innovative and

efficient solution to the traditional attendance-taking process. By leveraging advanced
facial recognition technology, the system automates attendance recording, ensuring
accuracy, reducing human intervention, and saving time. This approach eliminates issues
such as proxy attendance, manual errors, and delays associated with conventional
methods.
pg. 11
In conclusion, the project showcases how emerging technologies can revolutionize
mundane tasks, enhance operational efficiency, and provide a robust solution to
attendance management challenges. With further advancements and refinements, the
system holds significant potential for widespread adoption across industries and
educational institutions.

REFERENCES

[1]. Laranjo, L., et al. (2018). Conversational agents in healthcare: A systematic

review.Journal of the American Medical Association, 320(10), 1035-1047.
A comprehensive review of conversational agents in healthcare, covering their
uses and effectiveness.

[2]. Huang, Y., & Ngu, A. H. H. (2017). Medical chatbot systems: A review of
recent advances.Healthcare Informatics Research, 23(4),
331-337.
This paper reviews the advancements in medical chatbot systems and their
applications in healthcar

pg. 12