E-Pension System using DeepFace

Submitted in partial fulfillment of the requirements for the degree of

Master of Technology
in
Software engineering
by

DARRIN BRIGHT-22MIS0397
RAVANTH-22MIS0409
MEIGNANAVIGNESH-22MIS0364
HARIHARAN-22MIS0405

Under the guidance of

Dr.BALASUBRAMANI
SCORE,VIT, Vellore.

APRIL,2024

DECLARATION
We hereby declare that the thesis entitled “E-Pension System using DeepFace"
submitted by us, for the award of the degree of Master of Technology in
softwareengineering to VIT is a record of bonafide work carried out by me under the
supervision of Dr.Balasubramani.
We further declare that the work reported in this thesis has not been submitted and will not
be submitted, either in part or in full, for the award of any other degree or diploma in this
institute or any other institute or university.

Place : Vellore
Date :24.4.2024
Signature of the Candidate

CERTIFICATE

This is to certify that the thesis entitled “Thesis title” submitted by

Darrinbright,22MIS0397,Ravanth,22MIS0409,Meignanavignesh,22MIS0364,Hariha
ran,22MIS0405, Score, VIT, for the award of the degree of Master of Technology in
softwareengineering, is a record of bonafide work carried out by him / her under my
supervision during the period, 01. 12. 2018 to 30.04.2019, as per the VIT code of
academic and research ethics.

The contents of this report have not been submitted and will not be submitted either in
part or in full, for the award of any other degree or diploma in this institute or any other
institute or university. The thesis fulfills the requirements and regulations of the
University and in my opinion meets the necessary standards for submission.

Place : Vellore
Date : 24.4.2024
Signature
of the Guide

Internal Examiner External Examiner

Head of the Department

Programme

ACKNOWLEDGEMENTS
 Student Name

Executive Summary
Summary of the thesis
One page and not exceeding 200 words
Times New Roman, 12
CONTENTS Page
No.
Acknowledgement i

Executive Summary ii

Table of Contents Iii

List of Figures ix

List of Tables xiv

Abbreviations xvi

Symbols and Notations xix

1 INTRODUCTION 1

1.1 Objective 1

1.2 Motivation 2

1.3 Background 3

2 PROJECT DESCRIPTION AND GOALS 3

3 TECHNICAL SPECIFICATION 3

4 DESIGN APPROACH AND DETAILS .

4.1 Design Approach / Materials & Methods .

4.2 Codes and Standards .

.
4.3 Constraints, Alternatives and Tradeoffs
5 SCHEDULE, TASKS AND MILESTONES .

6 PROJECT DEMONSTRATION .

7 RESULT & DISCUSSION .

8 SUMMARY .

9 REFERENCES .

APPENDIX A .
INTRODUCTION
1.1. OBJECTIVE
The objective of integrating facial recognition technology, specifically utilizing the
DeepFace algorithm, within the E-Pension System is to significantly advance pension
management. By leveraging state-of-the-art technology, the system aims to ensure
enhanced security and user experience for pensioners.
Motivation:
The motivation behind this integration lies in the recognition of the potential benefits it
offers. DeepFace, known for its high accuracy in facial recognition tasks, serves as the
cornerstone of the system's biometric authentication process. This technology facilitates
swift and precise identification, thereby mitigating the risks associated with identity
fraud.
Background:
The integration of facial recognition technology within the E-Pension System not only
enhances security but also improves efficiency and user satisfaction. DeepFace's
advanced capabilities enable rapid and reliable authentication, streamlining the pension
administration process. This advancement signifies a departure from traditional
methods towards a more secure, transparent, and user-centric pension administration
process in the digital era.
TECHNICAL SPECIFICATION:
Key functionalities for e-Pension include robust user authentication, secure transaction
processing, and stringent data security measures. A comprehensive technology
selection process was conducted to identify the best facial recognition algorithms
technologies and platforms available in the market. The technology selection criteria
included accuracy, speed, scalability, compatibility with existing
infrastructure, and the most suitable facial recognition algorithm for integration. The
system design was then conducted to structure the e-Pension system, integrating the
best facial recognition technology. Data flow, system components, and interaction
between modules were defined to ensure efficient operation. Four primary components
work
together seamlessly to deliver a smooth user experience:
1. Frontend: The front end of the E-Pension System is built using HTML, CSS, and
JavaScript. It provides a user-friendly interface for pensioners to interact with the
system. This interface includes features such as login, account details, payment details,
pension details, etc.
2. Backend: The backend of the system is implemented using Django, a high-level
Python web framework. Django handles the business logic of the application, including
user authentication, database operations, and communication with the facial recognition
service.
3. Facial Recognition: This module is responsible for detecting the user's face using the
DeepFace algorithm and matching it with the image submitted by the user. It
communicates with the backend to validate the image of the user and proceed further.
4. Database: This uses an SQL database and stores the users details such as their
account details, personal details, pension records, and facial recognition data. This
module is accessed by the backend to retrieve and update any information
DESIGN APPROACH AND DETAILS:
Pension management is on the edge of a major transformation, ready to embrace the
potential of cutting-edge technology. The integration of facial recognition technology
into the E-Pension System is seen as
a beacon of innovation. As we go deep into the study, we uncover the intricacies and
reveal the seamless integration of technology and pension administration, leading to a
system poised to redefine security, efficiency, and user experience in pension
management. The main modules that make the E-Pension system are the Face Match
module, the Register module, and the Payment module. This section explores the
complex structure and functions of the key modules that make up the E-Pension
System:
Face Matcher module:
The Pensioner Face Matcher module is designed to identify and match faces in images
with a preloaded reference image database specifically tailored for pensioners. This
module primarily serves the purpose of identity verification or recognition, commonly
used in pension disbursement systems, or any scenario where confirming the identity of
pensioners is crucial. This module is designed to detect and match faces present in
target images with a preloaded database of reference images containing pensioners'
faces. Its main aim is to automate the identification process, enabling seamless
verification of pensioners' identities. By leveraging advanced face detection algorithms
and techniques, the module aims to accurately locate and
recognize pensioners' faces, thereby contributing to the efficiency and security of
pension-related systems. Face Detection utilizes advanced face detection algorithms
(OpenCV) to locate human faces within images or videoframes. These algorithms are
trained on large datasets todetect facial features and patterns robustly. Then the image
processing techniques are implemented to increase the quality and clarity of input
images, improving the accuracy of face detection and matching.This was constructed to
seamlessly help the pensioners to verify their identity. The module maintains a
reference image database containing facial images of registered pensioners. These
images are securely stored and indexed for efficient retrieval during the matching
process. This captures the real-time image and compares it with the reference image.
This algorithm for face detection is constructed using a deep face algorithm. This
constructed to seamlessly help the pensioners verify their identity This module was
integrated with the Django framework for my project using the views methodology.
Payment:
The payment gateway module tailored for pensioners was crafted to facilitate seamless
financial transactions within pension systems, specifically catering to the needs of
pensioners. This module addresses the necessity for a secure, user-friendly payment
mechanism tailored to the unique requirements and demographics of pension
beneficiaries. Its development is pivotal in ensuring timely
and efficient disbursement of pension funds, and welfare for pensioners.The module
employs a combination of secure payment protocols, encryption techniques, and API
integration with Razorpay's payment gateway services. This entails establishing secure
communication channels between pension systems Razorpay servers, implementing
tokenization for sensitive data, and adhering to
industry-standard security practices to safeguard pensioners'
financial transactions.The module is implemented using modern web development
technologies, with backend functionalities powered by
languages like Python and Django and frontend interfaces designed using HTML, CSS,
and JavaScript frameworks.Integration with Razor Pay is facilitated through their
API,leveraging secure authentication mechanisms and encrypted communication
protocols. Considerations during implementation include adherence to PCI DSS
compliance standards and scalability to handle varying transaction volumes. This
module was integrated with the Django framework using the views methodology
Register:
The Register Page module is designed to facilitate the registration of pensioners by
sending a one-time password (OTP) message to their registered mobile numbers for
verification. It prompts users to input their details, including personal information to
establish their identity and enable pension payments. Through a user-friendly interface,
the
module guides pensioners through the registration process, ensuring a smooth and
efficient experience.The module employs a robust authentication mechanism
bygenerating and sending OTP messages to pensioners registered mobile numbers.
Upon receiving the OTP, pensioners are prompted to enter it on the registration page
to verify their identity.The module is implemented using web development
technologies such as HTML, CSS, and JavaScript for the user interface. Server-side
scripting languages Python and
Django are used to handle form submissions and process
registration requests. Integration with a fast2SMS API
enables the generation and delivery of OTP messages to
pensioners' mobile numbers.By employing secure authentication mechanisms and user-
friendly interfaces, the module ensures accurate and efficient onboarding of pensioners
Continued refinement and enhancement of the module will further contribute to the
effectiveness and reliability of pensioner registration
processes. Continuous feedback integration from users ensures adaptive improvements,
aligning the module with evolving pensioner needs and industry standards.
PROJECT DEMONSTRATION:
fingerprint system which could be more reliable and efficient. The user can enter their
details on the account page. The pension payment is done using Razorpay which can be
seen If the users have any more doubts regarding any of the processes, they can reach
us through any one of the methods on the About page.

RESULT & DISCUSSION

The E-pension system has proved to be better than the traditional
fingerprint biometric in E-pension offices in many aspects
such as
1. Efficient: Face recognition is efficient because it scans the face of the pensioner
within a few seconds whereas in the traditional biometric, sometimes there is a high
possibility that the finger doesn’t scan the first time, it is required to
keep the thumb two or three times which can be annoying to some pensioners.
2. Secure: Facial recognition reinforces the safety of pensioners as the DeepFace
algorithm uses unique
mathematical patterns to store biometric data.
3. Physical Contact: The face recognition system avoids physical contact which is
better always.
The utilization of biometric authentication can be unreliable and inefficient sometimes.
This is where we propose an alternate methodology E-pension with DeepFace where
the
integration of facial recognition technology stands out withenhanced security and
efficiency in pension managementsystems. Facial recognition can enhance pension
systems
with secure and user-friendly identity verification methods.
However, certain challenges and considerations arise.Privacy is a primary concern,
raising questions regardingdata collection, storage, usage, accessibility, and security.
Clear regulations are crucial to ensure responsible utilization
of facial data, safeguarding privacy rights.Facial recognition technology, as outlined in
the proposed methodology, demonstrates advancements in accuracy,speed, and
scalability. The integration of deep learningalgorithms, such as the DeepFace
algorithm, enables
seamless verification of pensioners' identities, contributing to the efficiency and
security of pension-related systems. By analyzing various facial attributes such as the
size and shape of the eyes, nose, and mouth, facial recognition algorithms
can generate highly precise biometric templates for eachuser. It also offers versatility
and scalability, making it well-suited for deployment across diverse environments and
user demographics.
Whether pensioners access thee-Pension system through desktop computers or mobile
devices, facial recognition can provide seamless and consistent authentication
experiences.To get the most out of facial recognition technology inpension plans, it is
important to have clear rules andregulations governing its use. It is important to use
facial recognition technology that respects the privacy and rights of individuals,
avoiding any actions that may violate orabuse personal information By complying with
these principles, pension schemes can effectively use facial recognition technology to
improve safety and efficiency while adhering to ethical standards. By doing all of these
things, we can tap into the benefits of facial recognition technology while making sure
it's used responsibly and without any major issues

SUMMARY:
Successfully integrating facial recognition technology into pension planning
requires a comprehensive approach that addresses various factors. Firstly, establishing
clear rules and regulations to govern technology usage, ensuring compliance with
legal requirements, and safeguarding individual rights, including obtaining consent for
facial data collection and processing, is essential. Secondly,implementing careful
business practices involves conducting thorough risk assessments, implementing
mitigation measures, and regularly auditing the system for effectiveness
and compliance. Thirdly, upholding unwavering ethical standards prioritizes privacy,
data security, and transparency in technology usage, ensuring facial data is securely
storedand used only for its intended purpose, with clear communication about its
utilization. In conclusion,E-Pension with DeepFace represents a shift in user
authentication within e-Pension systems. Its combination of accuracy, convenience,
security, and scalability positions it as the preferred methodology for identity
verification in modern digital environments. By using facial recognition technology,e-
pension systems can increase security, simplify user experiences, and create a more
effective and user-friendly approach to E-pension management.

References
[1] “[1] Khin Sandar Win, EMBF II - 23, EMBF 6th Batch, 2018-2019,

PERCEPTION OF PENSIONERS ON E-PENSION SERVICE OF

MYANMA ECONOMIC BANK

[2] Hlib Filipchuk, Digital transformation in Ukraine, Ubezpieczenia

Społeczne. Teoria i praktyka nr 4/2021

[3] JURICA ŠIMURINA, Faculty of Economics and Business University of

Zagreb, DOMAGOJ HRUŠKA Faculty of Economics and Business

University of Zagreb, MILIVOJ MARKOVIĆ Faculty of Economics

and Business University of Zagreb

[4] Ei Khaing Zar Nwe, Roll No. MBF - 13, MBF- Day 1st Batch,

2017-2019, Yangon University of Economics

[5] João Brogueira de Sousa, Julián Díaz-Saavedra, Ramon Marimon, A

Worker's Backpack as an alternative to PAYG pension systems

[6] Dr. Alfred Ouma Shem, Head of Secretariat, Leonard Apiyo Audi,

Member,Charles Machira, Interim Chair,African Pension Supervisors

Association

[7] Behavior-based Cognitive Architecture for Meditative E-Learning,Hiran

Ekanayake, Damitha D karunaratna, K.P.Hewagamage

[8] The Image of Ukraine in the International Arena in the Minds of the

Current Processes of Digital Transformation and e-Government,Anna

Rudnieva1,Yuliia Malovana2

[9] Public Pension Provisioning for Old Persons in India, Kinjal

Sampat,Harsh Mander

[10] Pension Funds, the Requirement of Providing the Minimum

Guaranteed Return and Excessive Legislative Restrictions of Pension

Fund Investments, Tanja Markovič Hribernik, Igor Jakopanecs