Download ebooks file Machine Learning for Biometrics: Concepts, Algorithms and Applications (Cognitive Data Science in Sustainable Computing) Partha Pratim Sarangi all chapters
Download ebooks file Machine Learning for Biometrics: Concepts, Algorithms and Applications (Cognitive Data Science in Sustainable Computing) Partha Pratim Sarangi all chapters
com
OR CLICK BUTTON
DOWNLOAD NOW
https://ebookmass.com/product/machine-learning-for-signal-processing-
data-science-algorithms-and-computational-statistics-max-a-little/
ebookmass.com
https://ebookmass.com/product/introduction-to-algorithms-for-data-
mining-and-machine-learning-yang/
ebookmass.com
https://ebookmass.com/product/fundamentals-of-machine-learning-for-
predictive-data-analytics-algorithms/
ebookmass.com
https://ebookmass.com/product/introduction-to-algorithms-for-data-
mining-and-machine-learning-xin-she-yang/
ebookmass.com
Machine Learning for Business Analytics: Concepts,
Techniques and Applications in RapidMiner Galit Shmueli
https://ebookmass.com/product/machine-learning-for-business-analytics-
concepts-techniques-and-applications-in-rapidminer-galit-shmueli/
ebookmass.com
https://ebookmass.com/product/machine-learning-algorithms-for-signal-
and-image-processing-suman-lata-tripathi/
ebookmass.com
https://ebookmass.com/product/machine-learning-for-planetary-science-
joern-helbert/
ebookmass.com
Edited by
Madhumita Panda
Assistant Professor, Master in Computer Applications, Seemanta Engineering College,
Jharpokharia, Baripada, Odisha, India
Subhashree Mishra
Assistant Professor, School of Electronics Engineering, KIIT University, Bhubaneswar,
Odisha, India
Banshidhar Majhi
Director, IIITDM, Kancheepuram, Chennai, Tamil Nadu, India
Series Editor
Arun Kumar Sangaiah
School of Computing Science and Engineering,Vellore Institute of Technology (VIT),
Vellore, India
Academic Press is an imprint of Elsevier
125 London Wall, London EC2Y 5AS, United Kingdom
525 B Street, Suite 1650, San Diego, CA 92101, United States
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
Copyright © 2022 Elsevier Inc. All rights reserved
No part of this publication may be reproduced or transmitted in any form or by any means,
electronic or mechanical, including photocopying, recording, or any information storage and
retrieval system, without permission in writing from the publisher. Details on how to seek
permission, further information about the Publisher’s permissions policies and our arrangements
with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can
be found at our website: www.elsevier.com/permissions.
This book and the individual contributions contained in it are protected under copyright by the
Publisher (other than as may be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and experience
broaden our understanding, changes in research methods, professional practices, or medical
treatment may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in evaluating
and using any information, methods, compounds, or experiments described herein. In using such
information or methods they should be mindful of their own safety and the safety of others,
including parties for whom they have a professional responsibility.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors,
assume any liability for any injury and/or damage to persons or property as a matter of products
liability, negligence or otherwise, or from any use or operation of any methods, products,
instructions, or ideas contained in the material herein.
Numbers in paraentheses indicate the pages on which the authors’ contributions begin.
Sumitav Acharya (143), Department of Computer Science and Engineering, National
Institute of Science and Technology, Berhampur, India
A. Anandh (105), Department of Computer Science and Engineering, Kamaraj College
of Engineering and Technology, Madurai, India
Saurabh Bilgaiyan (155), School of Computer Engineering, KIIT Deemed to be
University, Bhubaneswar, Odisha, India
P.V.S.S.R. Chandra Mouli (65), Department of Computer Science, Central University
of Tamil Nadu, Thiruvarur, India
Manisha P. Dale (1), MES College of Engineering, Savitribai Phule Pune University,
Pune, India
Rupam Das (155), School of Electronics Engineering, KIIT, Deemed to be University,
Bhubaneswar, India
K. Devendran (87), Computer Science and Engineering, Kongu Engineering College,
Erode, Tamil Nadu, India
Sachit Dhamija (155), School of Electronics Engineering, KIIT, Deemed to be
University, Bhubaneswar, India
R. Jai Ganesh (129), K. Ramakrishnan College of Technology, Trichy, India
G.K. Kamalam (177), Department of Information Technology, Kongu Engineering
College, Erode, Tamil Nadu, India
Vaishali H. Kamble (1), AISSMS IOIT, Savitribai Phule Pune University, Pune, India
M. Kavitha (129), K. Ramakrishnan College of Technology, Trichy, India
P. Keerthika (87,177), Department of Computer Science and Engineering, Kongu
Engineering College, Erode, Tamil Nadu, India
Chirag Kyal (29), Department of Computer Science and Engineering, National Institute
of Science and Technology, Berhampur, India
K. Logeswaran (177), Department of Information Technology, Kongu Engineering
College, Erode, Tamil Nadu, India
V.M. Manikandan (201), Computer Science and Engineering, SRM University-AP,
Amaravati, Andhra Pradesh, India
R. Manjula Devi (87,177), Department of Computer Science and Engineering, Kongu
Engineering College, Erode, Tamil Nadu, India
xiii
xiv Contributors
xv
xvi Preface
The editors would like to gratefully acknowledge all of the contributors for con-
tinuous effort and timely submission of their chapters. This book would not
have been feasible without the cooperation of the chapter authors. All the chap-
ters have been reviewed for several rounds to facilitate the selection of final
chapters in our book. Valuable suggestions and guidance from the reviewers
helped the authors in refining individual chapters. Thanks also go to the
reviewers in enhancing the quality of the chapters of this book.
The editors would like to extend gratitude to Sonnini Ruiz Yura from
Elsevier for inspiration over the year. We would like to gratefully acknowledge
Andrae Akeh and Judith Clarisse Punzalan (Elsevier) for their patience during
the preparation of this book. In addition, we shall thank Swapna Srinivasan at
Elsevier for her sincere help and patience during the final preparation of this
book. Finally, we shall extend gratefulness to our family members and friends
for all their support.
xix
Chapter 1
1 Introduction
Face of an individual is a popular and well-accepted biometric trait that can be
used to perform identity recognition of adults as well as children. Children are
the most valuable and jeopardy group in society; hence they should be under
supervision continuously. Security and healthcare of children is an important
aspect of all countries [1]. Automatic recognition of children using their face
is a useful investigative tool to help identify missing children. Though the
development of the face of a child starts in the mother’s womb from 3 months,
it is not proportional to the development of other parts of the body. Therefore,
recognition of children below 6 years is still an open research problem. Children
recognition using different modalities needs to be studied to solve the problems
related to security and healthcare. As per the literature survey, still there is not
even one commercial biometric system in use for recognition of toddlers. Var-
ious researchers have discussed about biometric recognition of adults; however,
very few papers on toddlers are available. A meager amount of work has been
traced in biometric identification of toddlers or children. It is most challenging
to recognize a toddler from his own single photograph after a few months. But,
in some instances, such as missing children, we have only the face image. So,
the recognition of children from their face image is very important. Facial
images can be acquired without users’ active involvement using ordinary cam-
eras from a distance. Also in survey, we noted that that most of the toddler’s
biometric recognition is in verification mode.
Sahar Siddique studied longitudinal face recognition using the Extended
Newborn Database and Children Multimodal Biometric Database. Identifica-
tion accuracy achieved using CNN is 62.7% and 85.1% on both the databases,
respectively [2]. Rowden et al. studied the longitudinal face recognition of chil-
dren between 0 and 4 years of age. The same-session accuracy they achieved is
93% and cross-session accuracy 43% after 6 months [3]. Local Binary Pattern
(LBP) is widely used texture-based method for recognition of face biometrics
[4]. This method is also used for children face recognition [5,6].
This chapter is presenting a machine learning and deep learning approach
for face recognition of children. As per the literature review, the study of chil-
dren recognition with the help of their face modality started in 2010 in India
[5,6]. Most of the papers studied same-session face recognition. The meager
amount of work is carried out longitudinally, that is, images of the same subject
taken over a period of time. There are very few readily available databases for
newborns like FG-NET [7]. There are very few images of children below
5 years in these datasets. CMBD database of IIIT Delhi (India) [8] is one of
the children longitudinal face recognition databases, but due to security reasons
the face database is not publicly available [9]. Therefore, database collection is
the major task in infants and toddler’s recognition. For this study, the database
of 81 subjects for the same session and 48 subjects for the cross session is col-
lected. The span between the data acquisition sessions is 3–6 months. The major
stages of the proposed work are preprocessing of face images, which include
manually cropping of face images of size 120 120 and converting it into gray
scale. Feature extraction is based on principal component analysis and linear
discriminant analysis, and CNN is proposed. The classification of subjects is
done using machine learning classifiers on the children database of the same
session and cross session. Further, the work is extended using convolutional
neural network (CNN) in which we have proposed our own optimized model
with data augmentation used to compare the machine learning and deep learn-
ing classifiers on our database.
The major contributions of the proposed work are as follows.
1. Due to the nonavailability of reference databases for children below 6 years,
the collection of longitudinal databases of children faces itself is a great
challenge. In the proposed work, we have captured the face images of tod-
dlers with the mobile camera of resolution 20 MP, with consent from their
parents in two sessions. Time period between two sessions is 2 months to
1 year. In the same session, 730 images of 81 toddlers are taken. In the sec-
ond session, 485 images of 48 toddlers are captured.
2. In this chapter, we are proposing feature extraction using PCA, LDA, and
CNN approaches on the facial images and their comparative study. In PCA
and LDA approaches, we have applied different machine learning tech-
niques for classification of subjects, such as Support Vector Machine
(SVM), Logistic Regression (LR), Gaussian Naive Bayes, K-Nearest
Neighbors (KNN), Decision Tree, and Random Forest. In CNN, feature
extraction is done by convolution layers and classification is done by dense
layers.
Machine learning approach Chapter 1 3
54.8
27.8 80.5
307.4
Fingerprint Recognition
Facial Authentication
IRIS Recognition
Voice Recognition
Palm Recognition
Others
448.2
1679.4
useful in the case of missing children. Therefore, we are using face biometrics
for recognition of children. An example of different biometrics of children from
our database is shown in Fig. 2.
Human face is a 3D model, we can recognize it by its features such as eyes,
shape of face, and color. Automatic face recognition is based on 2D photo-
graphs of a person. In 1964 and 1965, Woody Bledsoe, along with Helen Chan
and Charles Bisson, recognized human faces using computers for the first time.
Nowadays adult face recognition achieves very high accuracy in the range of
99.63% [11]. Face recognition of children is still an open challenge. All face
recognition algorithms evaluate on false negative identification rate (FNIR)
which is dependent on age. The FNIR of children is higher as compared to
adults or seniors. The comparison of FNIR and false positive identification rate
(FPIR) is given in Table 1 for various stages of child age.
Child face development: Studies from the paper by Farkas discuss that in
the first 2 years of a child face development, mouth width of child increases
whereas mouth height decreases. Mouth shape alters from “rosebud-like” to
a more adult type. Growth of facial features is very fast in the first year, less
rapid in the second year. Subsequent changes were slow and irregular from
the age of 3 to 9 years [12].
I WAS awakened this morning with the chime which the Antwerp
Cathedral clock plays at half hours. The tune has been haunting me
ever since, as tunes will. You dress, eat, drink, walk, and talk to
yourself to their tune; their inaudible jingle accompanies you all day;
you read the sentences of the paper to their rhythm. I tried uncouthly
to imitate the tune to the ladies of the family at breakfast, and they
say it is “the shadow dance of Dinorah.” It may be so. I dimly
remember that my body was once present during the performance of
that opera, while my eyes were closed, and my intellectual faculties
dormant at the back of the box; howbeit, I have learned that shadow
dance from hearing it pealing up ever so high in the air at night,
morn, noon.
How pleasant to lie awake and listen to the cheery peal, while
the old city is asleep at midnight, or waking up rosy at sunrise, or
basking in noon, or swept by the scudding rain which drives in gusts
over the broad places, and the great shining river; or sparkling in
snow, which dresses up a hundred thousand masts, peaks, and
towers; or wrapped round with thunder—cloud canopies, before
which the white gables shine whiter; day and night the kind little
carillon plays its fantastic melodies overhead. The bells go on
ringing. Quot vivos vocant, mortuos plangunt, fulgura frangunt; so on
to the past and future tenses, and for how many nights, days, and
years! While the French were pitching their fulgura into Chassé’s
citadel, the bells went on ringing quite cheerfully. While the scaffolds
were up and guarded by Alva’s soldiery, and regiments of penitents,
blue, black, and grey, poured out of churches and convents, droning
their dirges, and marching to the place of the Hôtel de Ville, where
heretics and rebels were to meet their doom, the bells up yonder
were chanting at their appointed half hours and quarters, and rang
the mauvais quart d’heure for many a poor soul. This bell can see as
far away as the towers and dikes of Rotterdam. That one can call a
greeting to St. Ursula’s at Brussels, ind toss a recognition to that one
at the town hall of Oudenarde, and remember how, after a great
struggle there a hundred and fifty years ago, the whole plain was
covered with flying French chivalry—Burgundy, and Berri, and the
Chevalier of St. George flying like the rest. “What is your clamour
about Oudenarde?” says another bell (Bob Major this one must be).
“Be still thou querulous old clapper! I can see over to Hougoumont
and St. John. And about forty-five years since, I rang all through one
Sunday in June, when there was such a battle going on in the
cornfields there as none of you others ever heard tolled of. Yes, from
morning service until after vespers, the French and English were all
at it, ding-dong!” And then calls of business intervening, the bells
have to give up their private jangle, resume their professional duty,
and sing their hourly chorus out of Dinorah.
What a prodigious distance those bells can be heard! I was
awakened this morning to their tune, I say. I have been hearing it
constantly ever since. And this house whence I write, Murray says, is
two hundred and ten miles from Antwerp. And it is a week off; and
there is the bell still jangling its shadow dance out of Dinorah. An
audible shadow, you understand, and an invisible sound, but quite
distinct; and a plague take the tune!
CATHEDRAL OF ANTWERP
Who has not seen the church under the bell? Those lofty aisles,
those twilight chapels, that cumbersome pulpit with its huge
carvings, that wide grey pavement flecked with various light from the
jewelled windows, those famous pictures between the voluminous
columns over the altars which twinkle with their ornaments, their
votive little silver hearts, legs, limbs, their little guttering tapers, cups
of sham roses, and what not? I saw two regiments of little scholars
creeping in and forming square, each in its appointed place, under
the vast roof, and teachers presently coming to them. A stream of
light from the jewelled windows beams slanting down upon each little
squad of children, and the tall background of the church retires into a
greyer gloom. Pattering little feet of laggards arriving echo through
the great nave. They trot in and join their regiments, gathered under
the slanting sunbeams. What are they learning? Is it truth? Those
two grey ladies with their books in their hands in the midst of these
little people have no doubt of the truth of every word they have
printed under their eyes. Look, through the windows jewelled all over
with saints, the light comes streaming down from the sky, and
heaven’s own illuminations paint the book! A sweet, touching picture
indeed it is, that of the little children assembled in this immense
temple, which has endured for ages, and grave teachers bending
over them. Yes, the picture is very pretty of the children and their
teachers, and their book—but the text? Is it the truth, the only truth,
nothing but the truth? If I thought so, I would go and sit down on the
form cum parvulis, and learn the precious lesson with all my heart.
But I submit, an obstacle to conversions is the intrusion and
impertinence of that Swiss fellow with the baldric—the officer who
answers to the beadle of the British islands—and is pacing about the
church with an eye on the congregation. Now the boast of Catholics
is that their churches are open to all; but in certain places and
churches there are exceptions. At Rome I have been into St. Peter’s
at all hours: the doors are always open, the lamps are always
burning, the faithful are forever kneeling at one shrine or the other.
But at Antwerp it is not so. In the afternoon you can go to the church
and be civilly treated, but you must pay a franc at the side gate. In
the forenoon the doors are open, to be sure, and there is no one to
levy an entrance fee. I was standing ever so still, looking through the
great gates of the choir at the twinkling lights, and listening to the
distant chants of the priests performing the service, when a sweet
chorus from the organ-loft broke out behind me overhead, and I
turned round. My friend the drum-major ecclesiastic was down upon
me in a moment. “Do not turn your back to the altar during divine
service,” says he, in very intelligible English. I take the rebuke, and
turn a soft right-about face, and listen a while as the service
continues. See it I cannot, nor the altar and its ministrants. We are
separated from these by a great screen and closed gates of iron,
through which the lamps glitter and the chant comes by gusts only.
Seeing a score of children trotting down a side aisle, I think I may
follow them. I am tired of looking at that hideous old pulpit, with its
grotesque monsters and decorations. I slip off to the side aisle; but
my friend the drum-major is instantly after me—almost I thought he
was going to lay hands on me. “You mustn’t go there,” says he; “you
mustn’t disturb the service.” I was moving as quietly as might be, and
ten paces off there were twenty children kicking and chattering at
their ease. I point them out to the Swiss. “They come to pray,” says
he. “You don’t come to pray; you—” “When I come to pay,” says I, “I
am welcome,” and with this withering sarcasm I walk out of church in
a huff. I don’t envy the feelings of that beadle after receiving point
blank such a stroke of wit.
Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.
ebookmass.com