Machine Learning for Biometrics: Concepts, Algorithms and Applications (Cognitive Data Science in Sustainable Computing) 1st Edition - eBook PDF - Download the ebook today and experience the full content
Machine Learning for Biometrics: Concepts, Algorithms and Applications (Cognitive Data Science in Sustainable Computing) 1st Edition - eBook PDF - Download the ebook today and experience the full content
com
OR CLICK HERE
DOWLOAD EBOOK
https://ebookluna.com/download/introduction-to-algorithms-for-data-
mining-and-machine-learning-ebook-pdf/
ebookluna.com
https://ebookluna.com/product/ebook-pdf-machine-learning-refined-
foundations-algorithms-and-applications/
ebookluna.com
https://ebookluna.com/download/machine-learning-for-planetary-science-
ebook-pdf/
ebookluna.com
Artificial Intelligence and Machine Learning for EDGE
Computing 1st Edition Rajiv Pandey - eBook PDF
https://ebookluna.com/download/artificial-intelligence-and-machine-
learning-for-edge-computing-ebook-pdf/
ebookluna.com
ebookluna.com
https://ebookluna.com/download/advances-in-subsurface-data-analytics-
traditional-and-physics-based-machine-learning-ebook-pdf/
ebookluna.com
https://ebookluna.com/download/big-data-analytics-introduction-to-
hadoop-spark-and-machine-learning-ebook-pdf/
ebookluna.com
Machine Learning for Biometrics
Concepts, Algorithms, and Applications
Cognitive Data Science in Sustainable Computing
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].
Title: Inhibition
Language: English
Eden XI was three parsecs distant, near Algol. For the next ten
hours Saxon paced the marvelously equipped lifeboat and absorbed
data from the robot recorder. He stared at the hard crystal ache of
the stars and thought of the Director's son. He thought about the
shining cities of Earth, and about stasis.
Stasis meant—control.
It meant control of a billion people, a rigid planetary economy. It
meant the Assassination branch of the Corps. Assassination
(carefully contrived to appear accidental) took care of those few
malcontents who were either too smart or too stupid to sign up for
colonization. It meant a gradual weeding out of the unsane, the
power-mad, it meant learning the true meaning of sanity and peace
and racial brotherhood.
And it meant the stagnation of science, a thick film of dust gathering
on the textbooks of the military tactician, and warships rotting at
anchor. It meant the white spire of the Stasis Administration Center
at New Washington, and the words graven over the golden portals:
Know thyself, Man. Or die!
Was the dream worth it?
Or was Man doomed to die like a brawling ape, playing with
lightning?
Saxon could not answer.
Meanwhile the colonies had to be inhibited. One interplanetary war
could smash the fragile structure so painstakingly built over the last
few hundred years. This was the turning point, the final cross-roads
of Man's destiny.
Saxon smiled bleakly.
Ultimately there would be a colony they could neither inhibit or
destroy. The adaptive ultimate. That colony would be Man no longer,
but Homo Superior.
But by then, it wouldn't matter.
The lifeboat came in on the night side of Eden XI, and hung above
the blue mountains like a basking shark. Saxon checked his
coordinates. This had been the original landing site, almost two
hundred years ago. He switched the infra-view on maximum, and
began to cruise in widening spirals. These sixth generation hops
were usually routine. If nomadic, a few political shifts could help
warp the culture into a set pattern. A simple matter to play the
visiting deity, pick one warped psychotic, and invest him with power.
A dictatorship was by far the best way of inhibiting a young culture.
Agricultural city-states were almost as easy. Designate a particular
crop as sacred, kill the rotation program, impoverish the land,
introduce serfdom.
By dawn, Saxon found what he was looking for. A row of cleared
fields and a farmhouse. He reconnoitered a hundred miles farther
and frowned. There was no clump of dwellings, no sign of a village
trading community.
He brought the ship down in a forest three miles away from the
farmhouse and camouflaged it to look like a great mossy boulder. He
spent the entire morning testing the atmosphere and the soil with a
savage patience. In the early years of the Corps, virus mutations had
taken a fearful toll of intermittent spotters.
Finally he discarded his uniform and selected a pair of homespuns
from the ship's wardrobe locker. Under the homespuns reposed his
utility kit, a miniature arsenal.
Late that afternoon he emerged from the forest and stood at the
edge of the cleared fields, a weatherbeaten itinerant, obviously
willing to chop wood for a meal. Abruptly his jaw muscle twitched.
The scene was pastoral, perfect.
The man, plowing the south forty. The little girl, playing in the
shadow of the sleepy farmhouse.
But no beast pulled that plow. A giant of a man with power and
intelligence stamped on his bronze features pushed the plow by
hand, in a die-straight furrow.
The little girl was blonde and elfin. She wore sandals, her tunic was
brief and plain. She was playing follow-the-leader—
With a robot.
The robot was tall. The sun struck sparks from its steel carapace as
it lumbered after the girl. Saxon stood frozen as she came flying
towards him in a burst of tossing blond hair and laughter, as she saw
him and came to a dead halt.
"Hello," Saxon said. He tried to smile.
"Hello." Her inflection was slurred. After six generations, naturally.
Her blue eyes sparkled. "Foot-sore, stranger?"
The words had the cadence of a ritual greeting. Saxon stared at the
robot and said carefully, "Yes."
"He's only a primer model," she said, following his gaze. "Next year
when I'm twelve Father promised to install secondary circuits. My
name's Veena. What's yours?"
Saxon introduced himself, as the giant at the plow came forward. His
white smile was a benediction, his voice a lambent organ. "Welcome,
rover. Haven't seen one of you in months. I'm Lang. Agricultural
hobbyist. You'll stay?"
His tone was almost pleading. Saxon nodded inarticulately, followed
them towards the farmhouse. His hands were shaking.
The interior of the house was—dimensionless.
For a moment Saxon thought he was still outside. A silver brook
tinkled through the mossy carpet that was the floor. The south wall
was a golden vista of ripe wheat rippling in the warm breeze that
ruffled his hair. Birds twittered in the sun-flecked foliage overhead.
"Nice house," Saxon said numbly.
Lang's smile was different. "A bit pretentious, I'm afraid. Grandfather
built it right after the landing. We've been too lazy to do much
remodeling. A remarkable man, Grandfather."
That explained it, Saxon thought in relief. One titan in an infant
colony, warping it into a Utopian mold, passing on the heritage of his
genius. How long, he wondered coldly, before they built starships
and returned to demolish the Earth which had exiled them?
"It must be wonderful to be a rover," Veena said wistfully. "Lang, can
I go with him when he leaves?"
"You haven't completed Basic Ecology. Mentor's waiting for your
afternoon session."
Veena pouted and went outside to her robot. Lang grinned. "The
precocious brat's beginning to ask him questions he can't answer.
Soon I'll have to install a few more circuits."
Saxon shivered. Regardless of scientific attainment, any culture is
vulnerable to inhibition.
So said his agent's handbook.
Later he met Veena's mother, Merl, a handsome woman with calm
gray eyes who served them dinner by firelight. It was a good dinner.
These colonists seemed like good people. A shame they qualified for
inhibition.
Gently, Saxon began to probe.
In only six generations the colonists has scattered throughout the
entire hemisphere. Although the matrix of their culture seemed to be
the individual family unit, they lived according to whim. Some lived
in small communal groups. Some lived alone. Some, by choice, were
wanderers, rovers. They had science. Their philosophy seemed
nebulous, based on a benevolent ecology, brotherhood with all living
things.
Saxon frowned.
Six generations ago, the ecology on this world hardly had been
benevolent for man. This area of the continent had been a steaming
marsh, swarming with hungry saurians. Now it was all meadow and
forest.
Saxon said thoughtfully, "Have you ever felt the need for
organization? For a leader?"
He leaned back and waited for the seed to sprout. Two years ago on
Eden VIII, near Rigel, he had said the same thing to a sixth-
generation shaman, and it took scarcely a month for the shaman to
start an intra-tribal war.
But now the seed fell on sterile ground. Lang said, "I don't
understand. Any problem which cannot be solved at family level is
referred to the annual council."
"A leader." Saxon was patient. "One strong man to represent
everybody. To settle all problems as he sees fit?"
"Remember, Father?" Veena prodded. "Those arboreal cannibals
Grandfather used to mention? They had a nomadic tribal culture
based on brute strength."
Lang nodded somberly. "Good analogy. The most favorable
extrapolation indicated a racial life expectancy of only ten thousand
years. Their emotional stability index was nil, they would eventually
have destroyed themselves. The first generation decided it would be
more merciful to exterminate them. An unwise decision, I think."
He launched into a spirited ethnological discussion with Veena, and
Saxon sat, numbly.
They had no emotional insecurity to feed, no power-hunger. No herd
instinct to pervert, nothing to utilize as destruction potential.
No cultural weakness.
The room they gave him was small and comfortable. For a time he
lay on the sleeping hammock, considering the situation. He was
beginning to like them. That in itself, was dangerous.
The house was very still.
He got quietly out of the hammock and crept towards the door. He
had to get back to the lifeboat, to feed facts into the monitor.
One thing disturbed him.
According to his agent's handbook, family-group anarchies didn't
need inhibition.
He was halfway across the plowed field when Mentor's iron voice
said, "Good evening."
Moonfire glimmered on metal. The robot stood impassively before
him. Saxon said slowly, "I was just going for a walk."
"You are our guest: I shall walk with you."
"I prefer to walk alone."
"Guests prefer company. The house of Lang must observe the basic
amenities."
Was there a hint of sardonicism in Mentor's voice?
They walked along the furrows, man and robot. Saxon felt beneath
his shirt for the utility kit. He kept his voice level.
"Am I a prisoner?"
"You are a guest."
"Did Veena tell you I might try to escape?"
A pause, while relays clicked silently.
"That is classified information."
Saxon's fingers were steady as they touched his tiny blaster.
Benevolent anarchy indeed! He said carefully, "Do the colonists
resent their exile?"
Another pause. Mentor's voice was a flat drone. "The concept is
meaningless, the question invalid."
Like hell it is, thought Saxon, and fired.
A cold blue wash of energy illuminated the robot. For a moment
Saxon was blinded. When vision returned he saw Mentor standing
immobile, unscathed.
"Please go back to bed," the robot said.
Saxon went back to bed.
Do not allow the emotional charm of any culture, nor any individual
of that culture, to sway your inhibition appraisal.
In the narrow confines of the lifeboat he repeated the quotation
grimly. Good inhibition agents are inflexible. He was a good agent.
For almost an hour he fed data into the monitor tapes. Then he
touched a stud and closed his eyes, waiting for judgment.
"Agricultural family-group societies are normally stagnant," the
monitor droned. "Such cultures, regardless of technological level, do
not warrent inhibition of any type. Reference: twelfth generation
check on Eden V."
The room spun. Saxon whispered, "But they have cybernetics,
matter transmitters."
"Regardless of technological level." The monitor was adamant.
This was madness. Saxon wiped his forehead and said, "Assuming
geographical isolation no barrier to united group action in the event
of emergency."
"United action is incompatible with family-group."
"Assume and advise!"
Relays chattered. Abruptly the entire panel flashed crimson. The
monitor spoke one word.
"Annihilation."
Saxon referred to his Inhibition handbook. He had never annihilated
a culture before.
One hour later he went into the forest. Birds sang overhead. The
sun dappled him in light and shadow. He stalked a small furry
quadruped that squealed at him from a log and brought it down with
his sonic pistol.
Back in the lifeboat he watched the animal regain consciousness in
an air-tight tank, and very slowly he pulled a lever. A green vapor
rolled into the tank. The quadruped screamed. The green vapor fed.
It was the penultimate in sporedom, yet it was more than a spore. It
had virus characteristics, and its propagation rate was almost
mathematically impossible. There was no known defense, and once
used, the entire planet was forever untouchable. To Saxon's
knowledge it had been utilized only once on Eden I.
At dusk, he took the lifeboat up fifty miles. He released the spores in
a widening spiral, and finally jettisoned the tank. He went into an
orbit at ten thousand miles, and waited.
It would take approximately a week.
It was a long week. Saxon slept little. He paced the cabin. He looked
at the stars and thought about a blue-eyed waif with tears in her
voice, begging him to stay.
After a week the lifeboat came down at the edge of a grassy plain.
Saxon took a sample of the contaminated atmosphere to determine
propagation rate.
The atmosphere was pure.
Some freak of expansion. One uncontaminated spot in a hemisphere
of death.
He selected another location. Then another. That evening he close
the coordinates of his original landing site and tested the air again.
Finally he went outside the airlock. He breathed deeply, and the air
was fresh and sweet, it smelled of forest and cool streams and
evening dew. In the blue dusk birds twittered. A small marsupial
very much like a squirrel scampered to the safety of a tree and
scolded him.
Saxon began walking.
At the edge of the forest he saw the familiar plowed field. The
farmhouse was a friendly beacon in the twilight.
"Hello," Veena said. She stood at the edge of the forest. She was
smiling. "Welcome home, rover."
For the next few days Saxon was the perfect guest. He argued
philosophical abstractions with the family by firelight; by day he
hiked in the woods with Veena and listened to Mentor give her
lessons. He asked questions.
"Veena, do you know what a microorganism is?"
"Benevolent or malignant?"
"Malignant. A plague."
She pursed her lips. "Organic or cultural?"
"Organic of course."
"Bacteria." Veena shrugged. "Quite a few of the first generation died
immediately after the landing. Until they adapted. Until they
analyzed the basic metabolism of the planet's dominant life-forms,
and constructed a neutralizer."
"A neutralizer?"
"A protective shell of ionized particles," she said patiently, "keyed to
the individual body-chemistry."
"Classified information," Mentor droned.
Saxon licked his lips. "You mentioned cultural microorganisms?"
"Much more deadly. I call them that, but Lang says I'm being
semantically unsound. War, for example. Racial inferiority. To date
we haven't found a cure." She broke off, and her eyes were shining
wet.
"But you don't have wars," Saxon said.
"No."
"Then?"
"We have a—ghetto," the girl said slowly. "I can't tell you about it.
Perhaps soon—"
Abruptly she changed the subject.
Slowly, Saxon's defenses began to crumble.
To all intents he was now a member of Lang's household, Veena's
adopted big brother.
Big brother—or pet?
It did not really matter.
Updated editions will replace the previous one—the old editions will
be renamed.
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.
ebookluna.com