Conference Paper Title*

Praneeth Kumar Vemula Rohith Kumar Pawar Yashonath Bayyavarapu

School of C.S.E School of C.S.E School of C.S.E
Lovely Professional University Lovely Professional University Lovely Professional University
Phagwara, India Phagwara, India Phagwara, India
praneethvemulakumar3@gmail.com pawarrohith3@gmail.com abhibayyavarapu@gmail.com

Vinod Kumar Reddy Gorantla Ramesh Banavath Aditya Bharadwaj

School of C.S.E School of C.S.E School of C.S.E
Lovely Professional University Lovely Professional University Lovely Professional University
Phagwara, India Phagwara, India Phagwara, India
gorantlavinodkumarredy@gmail.com rameshpika13149@gmail.com adityathakur1619@gmail.com

Seema Gaba
Assistant Professor
School of CSE
LPU Phagwara , India
seemagaba1812@gmail.com

Abstract—Attendance marking is a task that occurs regularly B. Internet of Things (IoT)

in the day-to-day life of any organization. Attendance works as
a quantitative metric to measure the consistency of a candidate.
The Internet of Things (IoT) is a technological advancement
The traditional way of marking attendance is a bothersome, that is redefining connection by integrating sensors and soft-
lingering, and error-prone task. This task, as time progresses ware into common objects to collect and share data online. IoT
for any organization to deal with collective troubles the Smart improves productivity, efficiency, and convenience in a variety
attendance marking system aims to tackle all the hurdles. It is of ways, and also brings up issues with data management,
completely automated and diligent as it rules out human error
and lends more time to complete the task at hand. security, and privacy. IoT promises to change how we interact
Index Terms—Internet of Things (IoT), Machine Learning with technology and the environment around us as it develops
(ML), Smart Attendance Marking Systems(SAMS), OpenCV, further.
CNN.
C. Artificial Intelligence (AI) and Machine Learning (ML)
Artificial Intelligence (AI) and Machine Learning (ML) are
I. INTRODUCTION
two game-changing technologies that enable computers to
The purpose of this review article is to investigate and learn from data, see patterns, and make intelligent decisions
evaluate how IoT, AI, and ML are integrated into the creation without the need for explicit programming. AI encompasses a
of intelligent attendance marking systems. These solutions pro- larger spectrum of technologies than machine learning (ML),
vide a revolutionary method of tracking attendance by utilizing which focuses on techniques that help computers perform
the power of networked devices, sophisticated algorithms, and better based on experience.
data analytics. They guarantee improved accuracy, real-time II. LITERATURE REVIEW
monitoring, and insightful insights.
Examining several papers, got to know about different
A. Smart Attendance Marking System (SAMS) methodologies used to replace several traditional practices to
advance the approaches to the management for attendance
Smart Attendance Marking System transforms traditional
system. The collective study of the approaches used by mul-
attendance monitoring by combining IoT, AI, and ML. By
tiple papers assisted in comprehending on IoT, AI and ML
using RFID/NFC technology, IoT sensors, or face recognition,
based technologies and integration of them helps the system
it can automatically gather data as people enter the building. To
in solving the difficulties.
improve accuracy, AI systems examine this data and identify
Sarker et al [1] total hardware based C application for
trends and abnormalities. ML gradually improves the system,
system developed with the combined use of RFID, Raspberry
making it more flexible to various settings and user actions.
Pi, Motion sensor, monitoring system, Biometric fingerprint,
Administrators may optimize resource allocation and inter-
Arduino Mega, key- pad, and LCD will do a specific task
vention methods by accessing real-time attendance records
which is taking attendance with the architecture are password
and forecast insights. Through ongoing learning, the system
authentication, biometric fingerprint sensors, and RFID read-
develops, increasing productivity and decreasing manual work.
ers. The usage of passive RFID tags and Grove 125kHz RFID
To handle any issues, privacy and security measures must be
readers is highlighted as the report explains the relationship
strong. This integration promises simplified operations and in-
between these parts and the processing unit. It gives a thorough
sightful data-driven decision-making across several industries.
overview of the biometric fingerprint sensor enrollment and way to ascertain if people are present is to use decision logic
verification procedures while emphasizing the significance of that relies on majority vote. The findings of tests carried out
DSP chips and algorithms for precise person identification. It to gauge the viability of the suggested strategy are shown in
also describes how to authenticate using a password, which the section on assessment and testing. The research contrasts
statistical multiframes analysis with frame- based accuracy
guarantees safe access to the attendance tracking system.
measures using example video sequences from educational
The integration of serial port connection for data reception
contexts. The outcomes show a notable increase in attendance
from hardware devices—which guarantees real-time atten-
marking accuracy following a convergence time,
dance tracking—is explained in this study. It also covers the demonstrating the effectiveness of the suggested technique.
functionality and design of the user interface, highlighting Yadav et al [4] the study on existing attendance record-
its resilience against student abuse and instructor- ing approaches of RFID-based and biometric-based systems,
friendliness. The study includes studies on biometric highlights their advantages and limitations. It also discusses
fingerprint sensors, password-based systems, and RFID, the potential of using smartphones for attendance marking
highlighting their useful- ness in a range of applications that and hybrid approaches combining RFID and biometrics can
go beyond attendance tracking. It also looks at pertinent assist for paperless, easy execution. Cloud-reliability by
literature on software development techniques, identity Google App Engine, REST web service, web application built
verification methods, and system design, providing a using PHP, Google Cloud SQL, MQTT protocol. With an
foundation for the suggested smart attendance system. It emphasis on RFID and biometric-based systems, the literature
highlights the affordability, dependabil- ity, and efficacy of review section looks at various methods of recording
the suggested system, providing a viable way to improve attendance. It discusses the advantages and disadvantages of
attendance monitoring and accountability in the each method while highlighting the importance of biometric
educational field. It also suggests areas for growth and recognition in preventing imitation. The report also highlights
development of the smart attendance system prototype, the ways in which attendance systems are evolving, such as
outlining future research possibilities. the use of cellphones for the purpose of documenting
Arsenovic et al [2] as a result, a Deep Learning, Con- attendance. The section on system design and development
volutional Neural Network (CNN) based face recognition ar- contains a detailed explanation of the hardware components
chitecture with systems based on algorithms for machine and software architecture utilized in the creation of the Smart
learning and computer vision provides possibilities in search Attendance System. A comprehensive approach to system
of practical solutions to be flexible and reduce errors. the implementation is shown by the use of Google Cloud services
aspects are face embeddings: FaceNet CNN, SVM Classifier, for software development and Raspberry Pi, fingerprint
face landmarks and image placement, face detector: CNN scanner, keypad, and LCD screen as hardware components.
cascade. The findings make it possible for additional study to Overall, the research advances our understanding by
achieve even greater precision. A thorough explanation of the providing a description of an actual IoT implementation in the
development process, beginning with dataset preparation and field of education. This suggested Smart Attendance System
augmentation, is given in the techniques section. It presents a gives educational institutions an economical and effective
unique method of picture augmentation designed specifically approach to control student attendance by eliminating the
for face recognition tasks and justifies the use of a limited drawbacks of conventional attendance systems and proposing
dataset. The creation of the face recognition model is then an automated alternative.
covered in detail, including the stages involved in face Sawall et al [5] during Covid-19 to tackle several
detection, picture preprocessing, em- bedding formation, and challenges the executive team embraced blended learning.
classification. The practical application of the suggested The data is sent to the cloud using Wi-Fi technology, where it
approach is demonstrated by the discussion of the model’s is gathered. Students' Wi-Fi-enabled devices provide their
integration into an already-existing RFID- based attendance RSSI readings to a remote server, which collates the readings
system. Outcomes and conversation The created facial to determine which students are present and which are absent.
recognition system’s three-month assessment is shown in this So Smart Atten- dance Marking Systems (SAMS) using
section. The model’s efficacy is demonstrated by the wireless technology protocol Wi-Fi (Wireless Fidelity) makes
confusion matrix and accuracy analysis, which yielded an it easier to mark student attendance. A detailed description of
overall accuracy of 95.02. the attendance system’s design and functioning is provided in
Heng Lee et al [3] a Real- time facial recognition using the proposed system model. The technology uses
machine learning algorithms like deep learning with camera trilateration—using access points and students’ Wi-Fi
factor and data processing for the facial recognition for Smart equipped devices—to figure out where each student is in the
Attendance. Field-of-view plays a prominent role in the classroom. With client-side and server-side interactions to
camera factor an optimistic camera with good video provide smooth data gathering and verification, the
resolution and proper installation using deep learning model architecture is logically organized. Through empirical testing,
can sig- nificantly increase the accuracy and reliability. The the experiment section shows that the sug- gested attendance
suggested accuracy technique, which includes modules for method is accurate and feasible. Many controlled studies are
sampling, filtering, and decision- making, is explained in carried out to evaluate location accuracy and success rate.
depth in the materials and methods section. In order to The findings show promise, with high success rates and
minimize false positives and negatives, sliding window reasonably precise location estimates, confirming the
filtering and sample frequency are discussed in the study. One suggested system’s effectiveness. Ultimately, the study offers a
strong response to the problems with conventional attendance this choice.
tracking, especially in light of the COVID-19 epidemic. The difficultie
s this
suggested approach provides a smooth, sanitary, and effective interpreta
way to check student attendance by utilizing Wi-Fi technology tion
and already-existing classroom infrastructure. represent
Zhao et al [6] the zige network-based architecture atten- a smart
attendanc
dance system is integrated with RFID sensors. When the e system
number of readers and positioning performance are entered prototype
into the database, the simulation results are shown as the CDF which is
distribution. Once the student's ID has been correctly integrate
d by
scanned, the attendance will be instantly noted for them. technolog
In order to create intelligent classroom attendance systems, ies such
the study suggests fusing IoT technology with RFID and as
facial recognition technologies. The drawbacks of sensors
make at
conventional tech- niques are addressed by this novel ease for
strategy, which provides effective, real-time attendance managem
management solutions. The development and operation of ent and
face recognition and RFID- based attendance systems are reliable
system t
both thoroughly explained in this article. It improves the o record
readability and comprehension of the suggested solutions by the
providing technical specifications, architectural schematics, identity
and mathematical derivations. To assess how well the of the
student
suggested systems operate, the research runs simulation and this
experiments. It examines variables including user happiness, also
reaction speed, and positioning accuracy and provides makes
empirical data to back up the systems’ efficacy. the
managem
ent to
time
III. COMPARATIVE ANALYSIS saving.
Literature Survey 2 2017 Marko FaceTim Face Accordin NA
Arseno e – Deep recognition, g to this
S Year Autho Title Technolog Summar Future vic, Learning Deep study, the
. of rs y and y Scope Srdjan Based Learniing, typical
publ Dataset Sladoje Face Convolutio forms of
n
ic- Used vic, Recogniti nal Nueral attendanc
o
catio Andras on Network, e systems
n Anderl Attendan Deep will be
1 2016 Dhima Design RFID (Radio Examinin To a, ce Learning, replaced
n and Frequency gSeveral improve Darko System and training with
Kumar impleme Identification difficultie this work Stefano dataset and RFID
), Biometric vic. augmentati ones that
Sarker, ntation of Fingerprint
s in the in future
Nafize Smart traditiona with co- on need the
Sensor,
Ishtiaq Attendan Arduino l opt IoT and Five carrying
ue ce Mega, attendane with volunteers of proper
Hossai Manage Passwords, managem existing Photograph RFID
n, ment C# ent Systems. s. cards
Insan System Language. system then can together
Arafat Using The increased be seen with
Jamil. Multiple fingerprints remarkab in mobile GPS.
Step and RFID ly such as devices There are
Authentic tags need to making connecte drawback
ation be access of d to s to both
registered attendanc dedicated of the
is done e sheet to servers, attendanc
with an the GPRS & e system
input ID students lightweig methods
and enroll to sign ht indicated
the ID’s. the Protocol. above.
proxies As Workers
are taken, MQTT is could
faking light misplace
the weight the RFID
attendanc data card or
e records, transmiss the
and time ion tracking
consumin Protocol device, or
g, So to since its a another
overcome perfect person
 could attendanc
make the e
check in markings
their .
place. Raspberry An end-
Conseque 4 2019 Vikas Cloud
Pi 3, Finger to-end
It is
Yadav, Based possible
ntly, this print IoT
G. P. Smart to
may rely scanner, system
Bhole Attendan improve
on Google prototype
ce the
machine App Engine for
System system so
learning ,Google efficient
for that it
and Compute attendanc
Educatio will
computer Engine, e
nal automatic
vision, Google managem
Institutio ally
which Cloud SQL ent in
ns generate
may be education
a report
more al
on
adaptable institutio
defaulters
and ns is
and alert
error- presented
the
free. in this
appropria
study. It
3 2019 Kuan Statistical Multiframe The Future te
may be
Heng Multi- Statistics evaluatio research authority.
used in
Lee, frames For the n's will The
place of
Sanjay Accuracy purpose of findings concentra student
attendanc
V Methodol marking indicate te on app now
e sheets
Addica ogy For attendance, that employin simply
with a
m, Ilya Atten- accuracy because g various displays
more
Krylov dance and an of camera attendanc
dependab
, Sergei Marking accuracy students' settings e, but it
le and
Nosov, System tool based uncooper to may be
effective
Mee on ative enhance improved
attendanc
Sim OpenVINO position the low- to
e system.
Lai, were fluctuatio resolutio provide
It
Zhan created and n and n and them
illustrates
Qiang assessed. occlusion severe with vital
how the
Lee, The , the occlusion notificati
Internet
Chung evaluation traditiona circumsta ons
of Things
Shien was l frame- nces. As regarding
may be
Chai conducted based an tests, fee
used to
using a accuracy example, payments
intelligen
video measure one , results,
tly
sequence is unable camera and other
manage
from an to resolve has a related
an
internal the wider informati
education
test. attendanc field of on.
al
e view By
institutio
marking (FOV) utilising
n's
problem. and several
attendanc
The focuses prototypi
e system.
suggested on the ng
Attendan
statistics first few boards,
ce data is
After a rows of such as
easier to
converge students, the
get and
nce time while the Raspberr
utilize for
(i.e., 110 second y Pi Zero
various
seconds), camera or ESP
apps to
it has has a 32,
examine,
been smaller among
monitor,
demonstr FOV. others,
report,
ated that This the
and
using technique attendanc
trigger
multifra resolves e device
alerts
mes with the prototype
since it is
sliding problem may be
kept in
window of low- further
the cloud.
filtering resolutio condense
and n precisio d, hence
majority n. decreasin
voting g its cost.
improves It may be
the made
accuracy into a
of commerc
student ial
 product Gang Classroo Network ent of
with an Zhao, m RFID college
attractive Meiho Attendan technology, education
appearan ng Qu. ce Face managem
ce for use Manage Recognitio ent, proce
as a ment n ss of
portable System technology building
device to Based on and RFID smart
take Internet and Face classroo
attendanc of Things recognition ms, to
e in database deal with
classes. from the issues
correspondi related to
5 2021 Emily COVID- Received The NA
ng author. regular
Sawall, 19 Zero- Signal studies
attendanc
Amber Interactio Strength demonstr
e, such as
Honnef n School Indicator, ate that
doing
, Attendan Wi-Fi the
repeated
Moha ce enabled suggested
roll calls
med System Devices and attendanc
and
Moha There was a e system
random
med,Al smart is a zero-
roll calls
i gadget effort,
in class
Abdull everywhere zero-
based on
ah S. in the 18 by interactio
the
AlQaht 9-foot n, and
circumsta
ani, efficient
space. nces to
Thamr Within the way to
keep an
aa space, the track
eye on
Alshay attendanc
access students
eb. points were e in a
and their
arranged in post-
tardiness,
a triangle. pandemic
early
For the society.
departure
experiment, Based on
, and
the access the
absenteei
points' experime
sm.
locations nt
. As a
were findings,
result,
measured. we think
this
that our
suggests
suggested
an
attendanc
intelligen
e system
t
is a good
attendanc
approach
e
because
managem
of its
ent
95.24%
system
success
based on
rate. The
RFID
trials
technolog
were
y, which
conducte
can
d at
precisely
various
detect
times of
student
the day,
absences
and we
and
think that
substituti
this
ons as
approach
well as
would
properly
motivate
record
teachers
and
to collect
identify
attendanc
student
e in a
arrivals
secure
and exits
and
from the
reliable
classroo
manner.
m.

zige’S 3G In
6 2022 Mingta College NA
Sensor developm
o Zhao, Smart
 IV. CONCLUSION [10] Rakhi Joshi, V. V. Shete, S. B. Somani, “Android Based Smart Learning
and Attendance Management System”, International Journal of Ad-
Using IoT (Internet of Things) and Machine Learnings(ML) vanced Research in Computer and Communication Engineering Vol. 4,
algorithms leverages the Smart Attendance Marking System Issue 6, June 2015.
(SAMS). At first the systems were introduced with individual [11] Asri Nuhi, Agon Memeti, Florinda Imeri, Betim Cico, “Smart At-
technology based on hardware running prototype monitored tendance System using QR Code”, 2020 9th MEDITERRANEAN
CONFERENCE ON EMBEDDED COMPUTING (MECO), 8-11 JUNE
with the software in computer system similarly to reference 2020.
[1]. The sensors, such as Biometric and biometric reading [12] Meng Zhi and Manmeet Mahinderjit Singh, “RFID-Enabled Smart
system, embedded with microcontroller Raspberrypi 3 which Attendance Management System”, Future Information Technology - II,
Lecture Notes in Electrical Engineering 329, 2015.
collects and analyzes the vast amount of data from the real [13] Rajarshi Samaddar, Aikyam Ghosh, Sounak DeySarkar, Mainak Das,
time environment along with Google cloud which improves Avijit Chakrabarty, “IoT Cloud based Smart Attendance Management
the machine as in instance [4]. Convolutional Neural System using RFID”, International Research Journal on Advanced
Networks (CNN), Deep Learning, OpenCV, and computer Science Hub Vol.05,Issue 03 March 2023.
[14] Raghav Apoorv, Puja Mathur, “Smart Attendance Management using
vision are examples of machine learning (ML) algorithms that Bluetooth Low Energy and Android”, IEEE Region 10 Conference 2016.
aid in facial identification of individuals and have [15] Xiong Wei, Anupam Manori, Nandgopal Devnath, Nitin Pasi, and Vivek
demonstrated promising outcomes in [2] and [3]. The Face Kumar, “QR Code Based Smart Attendance System”, International
Journal of Smart Business and Technology Vol. 5, No.1, 2017.
recognition technology integrated with IoT for excess [16] Unnati Koppikar, Shobha Hiremath, Akshata Shiralkar , Akshata Rajoor,
Authentication improved Smart Attendance marking system V. P. Baligar, “IoT based Smart Attendance Monitoring System using
(SAMS) has significantly increased the accuracy, re- liability RFID ”, 2019 1st International Conference on Advances in Information
and efficiency of the system tested in [4] and [6]. IoT Technology (ICAIT).
[17] Oluwagbemiga O. Shoewu1, Makanjuola N.T, Akinyemi L.A, “Smart
collects useful data from the real time environment with the Attendance Management System (SAMSYS) for an Academic Institu-
help of the existing technology and autonomous flow of data tion ”, ITSPOA journal volume 2, 2018.
between the devices. The Smart Attendance Marking [18] Khawla Alhanaee, Mitha Alhammadi, Nahla Almenhali, Maad Shatnawi,
“Face Recognition Smart Attendance System using Deep Transfer
System's intelligence and functionality are significantly Learning”, 25th International Conference on Knowledge-Based and
influenced by the application of machine learning (ML) Intelligent Information Engineering Systems, 2021.
methods. Calculate the required time in IoT data using
methods such as time series analysis. The integration of the
IoT and Machine Learning (ML) provides the real time data
analysis, decision making and Representation of the complex
data. The Smart Attendance Marking System (SAMS) with
extent of technology eliminates the management troubles
and makes this daily task hassle free and the best
replacement for the Traditional Attendance System.

REFERENCES
[1] Dhiman Kumar Sarker, Nafize Ishtiaque Hossain, Insan Arafat Jamil,
”Design and implementation of Smart Attendance Management System
Using Multiple Step Authentication”, 2016 International Workshop on
Computational Intelligence (IWCI), 12-13 December 2016.
[2] Marko Arsenovic, Srdjan Sladojevic, Andras Anderla, Darko Stefanovic.
”FaceTime – Deep Learning Based Face Recognition Attendance Sys-
tem”, SISY 2017 • IEEE 15th International Symposium on Intelligent
Systems and Informatics, Subotica, Serbia, September 14-16, 2017.
[3] Kuan Heng Lee;Sanjay V Addicam;Ilya Krylov;Sergei Nosov;Mee Sim
Lai;Zhan Qiang Lee;Chung Shien Chai, ”Statistical Multiframes Accu-
racy Methodology For Attendance Marking System”, 2019.
[4] Vikas Yadav, G. P. Bhole, ” Cloud Based Smart Attendance System
for Educational Institutions”, 2019 International Conference on Machine
Learning, Big Data, Cloud and Parallel Computing (Com-IT-Con), India,
14th -16th Feb 2019.
[5] Emily Sawall, Amber Honnef, Mohamed Mohamed,Ali Abdullah S.
AlQahtani, Thamraa Alshayeb, ”COVID-19 Zero-Interaction School
Attendance System”, IOT Electronics and Mechatronics Conference
(IEMTRONICS), 2021.
[6] Mingtao Zhao , Gang Zhao , Meihong Qu, ”College Smart Classroom
Attendance Management System Based on Internet of Things”, Compu-
tational Intelligence and Neuroscience Volume 2022.
[7] Kaneez Laila Bhatti, Laraib Mughal, Faheem Yar Khuhawar, Sheeraz
Ahmed Memon, “Smart Attendance Management System Using Face
Recognition”, EAI Endorsed Transactions on Creative Technologies, 27
August 2018.
[8] Syam Kakarla, Priyaranjan Gangula, M.Sai Rahul, C. Sai Charan Singh,
T. Hitendra Sarma, “Smart Attendance Management System Based
on Face Recognition Using CNN”, IEEE-HYDCON 2020.
[9] Shreyak Sawhney, Karan Kacker ,Samyak Jain, Shailendra Narayan
Singh , Rakesh Garg, “Real-Time Smart Attendance System using
Face Recognition Techniques”, 9th International Conference on Cloud
Computing, Data Science Engineering, 2019.
[19] Bruno Zoric´; Mario Dudjak; Drazˇen Bajer; Goran Martinovic´,”Design
and development of a smart attendance management system with Blue-
tooth low energy beacons”, 2019 Zooming Innovation in Consumer
Technologies Conference (ZINC).
[20] Chung Seng Keau, Chin Kim On, Mohd Hanafi Ahmad Hijazi,”Smart-
Hadir – Mobile Based Attendance Management System”, iJIM – Vol. 15,
No. 14, 2021.