An Investigation towards Effectiveness of Present State of Biometric-Based Authentication System

International Journal of Research and Scientific Innovation (IJRSI) | Volume IV, Issue VII, July 2017 | ISSN 2321–2705
www.rsisinternational.org Page 44
An Investigation towards Effectiveness of Present
State of Biometric-Based Authentication System
Gavisiddappa1
, Dr. M Shivakumar2
, Dr. Chandrashekar M Patil3
1
Associate Professor, Dept. of ECE, CIT, Gubbi, Karnataka, India
2
Professor & Head, Dept. of E&IE GSSSIETW, Mysore, Karnataka, India
3
Professor, Dept. of ECE, VVCE, Mysore, Karnataka, India
Abstract— The adoption of biometric-based authentication
mechanism has been already initiated a decade back but still in
real-life we get to see usage of only unimodal biometrics. Out of
all the different forms of biometrics, we see usage of fingerprint
as the dominant attribute in contrast to different other attributes
e.g. teeth image, palm, facial geometry, retina network, iris, etc.
Multimodal biometrics is believed to offered better security
compared to unimodal. Although, there are some of the technical
advancement in evolving up with new multimodal methodologies,
but still commercial usage of such is yet to be seen. Therefore,
this manuscripts aims to explore the level of effectiveness in
existing approaches of biometric-based authentication system in
order to further investigate the unaddressed solution towards this
problem. This paper reviews the approaches used for addressing
different problems associated with biometrics and discusses
about their technical methodologies as well as their limitations.
Keywords- Biometric, Authentication, Security, Unimodal,
Multimodal, Fingerprint, Face, Algorithms
I. INTRODUCTION
sage of biometric is one of the increasing mechanisms of
performing authentication from small to bigger scale of
application environment [1]. According to one recent article of
Forbes, it was mentioned that 50% of the smartphone
application will adopt using fingerprint sensor by the year 2019
[2]. The major forms of biometrics includes biological signals
e.g. fingerprint, retina structure, facial image, teeth image, etc,
which is very unique and discrete for one person. These unique
data is used as a digital identity of any person utilized for their
authentication. Inspite of advantages of biometrics in contrast
to conventional password, their implementation is no 100%
resilient against attacks [3]. Usage of biometrics allows
freedom from memorizing or storing any form of password for
accessing. One of the potential challenges of using biometric is
to ensure security against integrity and safety of biometric data.
There are various forms of unaddressed problems associated
with robust and effective usage of biometric-based
authentication system. The first concern about usage of
biometric is variations of biometric signal within a person. It is
highly possible that a biometric signal may differ from the
signal that is stored in biometric system over a period of time
due to many reasons e.g. stress, growing age, sickness, etc.
This results in higher degree of false positive alarms where
legitimate person could be denied access of their own accounts.
The second problem in biometric usage is associated with the
type of sensors used for scanning. The sensitivity, calibration,
age, embedded interface may vary with the surrounding
environment even if the biometric signals are unchanged. Such
system causes false positive while performing authentication.
The third problem is associated with type of feature extraction
algorithms. It is infeasible to perform matching of queried
biometric signal with millions of biometric signals stored,
hence, features are extracted which minimizes the search and
offer faster results. However, different feature extraction
algorithms have different performance which cannot be
standardized. If the mechanism for comparison scoring differs
from matching algorithm then the performance is seriously
affected [4]. The fourth problem is associated with the data
integrity which means that it is possible for a legitimate user to
corrupt/distort the data due to certain security loopholes.
Hence, existing system doesn’t offer much data integrity
towards biometric template security. The fifth problem is
associated with the higher degree of instability associated with
biometric signals. None of the biometric attributes offer 100%
stability either over shorter or over longer period of time
causing greater dependency of updating the template. All the
above mentioned challenges adversely affect the decision
making operation which completely works on probability. This
will mean that all the positive matching biometric signals will
represent a probability of precise identification while a negative
matching biometric signals represents the same probability an
not a concrete inference of conclusion. Therefore, there is
greater deal of challenges encountered by the usage of
biometric-based authentication system.
This paper reviews the existing approaches used by
researches for addressing the challenges associated with
biometrics emphasizing its usage in authentication system.
Section II discusses about the usage of the biometrics in the
form of fundamental discussion followed by existing system
.Section III with respect to approaches used and limitation.
Section IV discusses about the modern techniques used for
authentication by combining different forms of biometric
signals. Brief discussion of research gap is carried out in
Section V followed by proposed study contribution .In Section
VI in the form of conclusion. A brief discussion of future work
is also carried out in Section VI at the same time.
II. A SNAPSHOT ON BIOMETRICS USAGE
Biometric systems utilize an individual’s human body
characteristics, such as the face, iris, fingerprints and hand
prints which do not normally change over time and
U

identification involving the enrollment of these traits in a
database for future recognition purposes. The evolution of the
usage of the biometrics dates back on 1858 where images from
hand and finger were used for identification. Table 1
highlights that since 1858, the usage of fingerprint based
authentication system has become quite famous; however it
was more used for solving classification problems and less
into authentication. The next forms of modalities found to be
frequently used are facial image and retina image (i.e. iris).
These forms of biometrics are found to offer higher degree of
security and hence they were used for designing security-
based application that demands resilient authentication
mechanism.
Table 1 Evolution of Usage of Biometrics [5]
Year Modalities Purposes Description
1858 Hand and Finger image Identification First time the image of the palm was recorded
1870 Anthropometries Identity individuals Capable of recording the physical measurements.
1892 Finger print Categorizations Used for preliminary identification of users
1894 Finger print Categorizations Using fingerprint to solve crime-based investigation
1896 Finger print Categorizations Using fingerprint to solve crime-based investigation
1903 Finger print Identification For the identification of criminals.
1936 Iris Identification Iris patterns as a method to recognize an individual.
1960s Face Reorganization
It calculated distances and ratios to a common reference
point that was compared to the reference data.
1960 Speech Speaker Reorganization
It is based on the analysis of x-rays of individuals making
specified phonic sounds.
1965 Signature Recognition Automated Signature Recognition
1969 Finger print Automatic Recognition
Challenges: Scanning, Comparing and Matching Finger
prints in a minute.
1970s Face Automatic Face Recognition
Used 21 specific subjective markers such as hair color and
lip thickness to automate face recognition.
1970 Speech Complex Behavior Used motion x-ray and included tongue and jaw.
1974 Hand Geometry Recognition
Physical access, time, attendance and personal
identification.
1975 Finger print
Recognition and digital
storage
Reduce the cost and digital storage
1976 Voice Speaker recognition Speaker recognition
1977 Signature Personal Identification Used in US air force for authentication
1980s Speech Speaker Recognition Can assists in recognizing voice of speaker
1985 Two Iris Similarity measurement
Can perform similarity matching between two different
form of iris.
1985 Hand Geometry Identification
Used for authentication in banking system using hand
geometry in US.
1986 Finger Print
Exchange of Finger Print
Minutiae data.
Considered as an international standard.
1986 Iris Identification
Can perform involuntary recognition of iris pattern of
human
1988 Face-Video images Recognition System Carried out exhaustive database search
1988 Face Recognition System
Offer magnified version of multiple images of facial
recognition
1991 Face Recognition System
Offers supportability of involuntary recognition of facial
image.
1993 Face Face recognition
Assessment of facial validation from simple to complex
mode of facial orientation
1994 Iris Iris Recognition First iris recognition system.
1994 Finger print Identification Involuntary fingerprint identification
1995 Iris Recognition system Commercial Product
1996 Hand geometry Identification To manage & safeguard user’s entry to specific location
1996 Speech Recognition System Targets enhancing speech recognition performance
1998 DNA
To digitally store, search and
retrieve
Used for performing DNA sequencing for target suspect
identification
1999 Finger print Identification Storage of digital image, supports exchanging of data
2000 Finger print, iris Identification Recognition using vascular patterns
2002 Standardization of biometrics by ISO/IEC
2003 Establishment of Forum for European Biometrics
2008 Initiated coordination of biometric database by US government
2010 Adoption of biometric antiterrorism activity by US National Security
2013 Inclusion of biometrics in smartphone

In the last two decades, the concept of biometrics is associated
with huge amount of management systems that increases the
open challenges between individual security as well as
enterprise security. Biometric techniques are growing popular
in public security related applications, such as customs
control, building entrance control and terrorist identifications
[6]. The procedure of identifying a person using security
device is named as authentication device. A biometric system
is the pattern recognition system that is used for extracting a
feature set from the data, and comparing the feature set in the
data base.
Figure 1 Classification of different Biometrics
Multimodal biometric devices are being progressively more
installed in various biometric applications as they overcome
limitations of unimodal biometrics. Based on the application
scenario, a biometric device may work either in identification
mode or a verification mode. Fig.1 shows the classification of
different biometrics system. Biometrics is mainly classified
into physical and behavioral systems. These are further
classified into head, hand, face, iris, fingerprint, hand
geometry, palm print, signature, voice etc [7] [8]. The
conventional means of using biometric authentication system
consists of extracting the feature from the different forms of
biometric signals and generate a template. This phase is called
as enrollment phase which is followed by template matching
phase. In the template matching phase, the queried (or input)
of new biometric signal is subjected to feature extraction
which is followed by template matching phase called as
verification. Fig.2 showcases the process of enrollment and
verification in existing system.
Figure 2 Existing Biometric Authentication
Although biometrics are increasingly used in authentication
system in majority of the commercial places, but it suffers
from some significant challenges e.g. i) privacy issues, ii)
security issues, and iii) irreplaceability issue. The
conventional biometric-based authentication system uses
central entity in order governs the complete private
information of the user. Such central entity is always assumed
not to be compromised and that is where it leads to serious
privacy issues. The security issues arise when such biometric
template has higher level of correlation of user’s information
with each other. Such information within the template
becomes the prime point of attention for the adversaries. Once
the biometric template is compromised that collateral attacks
begins. The third problem is related to the replication of the
biometric details in order to spoof the original one. Table 1(a)
showcases the scale of effectiveness of different forms of
biometric systems that are used in authentication system till
date.
Table 1(a) Existing Factors of biometric system [1]
Factor Effectivity Scores of Various Biometric System
Finger
Print
Face
Hand
Geometry
Iris Voice Signature Gait Ear
Hand
Vein
Palm
print
Accuracy H L M H M M H H H H
Ease of Use H M H M H H M M M M
Cost L M M H H L H H H H
Privacy H H M H H H H L L M
Distinctiveness H L M H L M M H H H
Universality M H M H M L M M M M
Performance H L M H L L L M M H
Collectability M H H M M H H M M M
Acceptability M H M L H H H H M M

III. EXISTING TECHNIQUES
This section discusses about the some of the existing
techniques implemented during the year 2010-2017
emphasizing on the enhancing the features of biometric system
for further trustworthy authentication system. There are
various ranges of the studies being presented with different
methodology
A study towards usage of biometric for securing the peer-to-
peer payment process over the network is introduced most
recently by Ghosh et al. [9]. The authors have discussed a
prototype model where mobile phone can be used for
extracting the finger print information along with account
number of device. The transaction of the information uses
common GSM network for authentication. However, the
technique highly depends on the device ID as well as user ID,
which is easier to be faked if the device is stolen. Usage of
biometric was also seen to be used in wireless sensor network
as seen in the study of Choi et al. [10]. Fuzzy logic was used
for strengthening the security features to be used for
authentications with an aid of cryptography-based approach.
The technique uses a mutual authentication scheme in order to
resist guessing & replay attack however the technique doesn’t
offer enough claims to show its practical usability. Different
level of biometric was also found in many studies. The work
done by Choi et al. [11] have used electrocardiogram as the
biometric attribute used for authentication. Such forms of bio-
signals are generated by the sensors residing in wearable
devices. The technique uses radial-basis function as well as
supervised learning for performing an effective classification
used for authentication. Studies towards strengthening
biometric-based authentication system was also carried out by
Bhuiyan et al. [12] where the retinal-vessel was used as an
biometric attribute for performing authentication. The
technique extracts the unique vascular network of retinal and
performs feature extraction which is further subjected to
geometric hashing method. The method finally extracts certain
invariant feature, which otherwise is challenging to be
extracted by conventional biometrics. The study outcome has
proved to offer 96% of precision rate. While working towards
biometrics, the user’s identity is at a great stake of privacy.
Hence, existing literatures have also implemented techniques
towards retaining maximum privacy while performing
biometric-based authentication.
The work carried out by Sedenka et al. [13] has formulated a
model for ensuring highest privacy using Principal Component
Analysis (PCA). The technique has enhanced protocol for
Garbled circuits and applied Advanced Encryption Standard
(AES) for further encryption. Kang et al. [14] have presented a
study where the resiliency of existing biometrics can be
evaluated. The author uses finger print mainly to perform
authentication with an aid of a sophisticated mathematical
modeling. Liu et al. [15] have presented a discussion of
enhancing the system performance during biometric
authentication. The technique presents discussion of linear
fusion to achieve robustness. The work carried out by
Venugopalan et al. [16] have discussed about usage of heart
beats to be used as biometric authentication. The authors have
also presented an analysis of such biomedical waveforms
using impedance factor. Jiping et al. [17] have presented a
cryptographic-based authentication scheme in biometrics.
Frank et al. [18] have presented a model called as touch altaics
that uses pattern analysis of the touch behaviour of the user on
the screen as a biometric trait for performing authentication.
Klonovs et al. [19] have presented a discussion of
authentication system using Electroencephalogram. On the
other hand, biometric template plays a critical role in every
biometric based authentication that is required to be
safeguarded. Studies in such direction have been carried out
by Simoens et al. [20] where multiple complex attacks have
been studied. The study carried out by Biggio et al. [21] has
analyzed the impact of practical adversary associated with
spoofing on authentication system in biometrics. Considering
the case of facial and fingerprint expression, the authors have
presented an analysis for both single and multimodal based
authentication. Urtiga and Moreno [22] have explored another
novel mechanism of biometrics using keystrokes that analyzes
the typing pattern of user. Sui and Du [23] introduced the
concept of biometric capsule where it computes the difference
between the reference subjects. The technique extracts
significant features of user followed by generation of bio-
information to further recollect more unique feature. The
complete process generates significant feature called as bio-
capsules that are stored and used for next step of verification.
Safie et al. [24] have presented a technique where the
heartbeat is used for biometric signal. The technique
introduces a new form of feature that is extracted from active
pulse ratio. Kumar and Zhang [25] have presented a technique
where different scores of matching is generated from different
templates of the user. The technique uses palm print as well as
shape of the hand as biometric signal to carry out
authentication. Chatterjee et al. [26] have used signal
information from the postures for performing biometric-based
authentication along with usage of radial-basis function for
further strengthening the authentication system. Tao and
Veldhuis [27] have presented facial attribute as biometric for
authenticating in mobile device. Upmanyu et al. [28] have
presented a technique called as blind authentication system
that is meant for safeguarding the biometric template. The
technique is designed using public key cryptography for
offering flexible encryption to the biometric template.
Hence, it can be seen that there are various studies targeting
at enhancing the performance of the biometric-based
authentication system. The existing studies are much inclined
towards evolving up with a new techniques and methodologies
to strengthening the biometric-based attributes thereby
assisting resiliency against multiple forms of adversaries.
Table 2 outlines the existing research contribution.

Table.2 Scaling Effectiveness of Existing techniques of Biometric-based Authentication (B: Biometric, M: Methodology)
Authors Problem Technique Advantage Limitation
Ghosh [9] Secure biometrics B: Fingerprint, Device ID
M: Experimental
Easier peer-to-peer transfer Possibility of device ID to be
compromised is not considered
Choi [10]. Securing biometric
authentication
B: hypothetical biometric
imprints
M: Analytical, mutual key
exchange
Resist multiple form of attack Few extensive analyses to
prove/benchmark the outcome.
Choi [11] Adoption of
physiological sensor
B: Hear-beat
M: Experimental, Support vector
Machine, radial basis function
Higher authentication
performance
Induces delay due to learning
mechanism involved.
Bhuiyan [12] Biometric
authentication
B: Retinal vessel
M: analytical, feature extraction
using geometric hashing
Higher recall and precision Success rate depends on threshold
value between 20-30.
Sedenka [13] Privacy preservation B: hypothetical biometric
imprints
M: Analytical, PCA, AES
Light-weight protocol No extensive analyses to
Kang [14] Resiliency of
biometrics
B: fingerprint/ DNA
M: Mathematical
Mathematically sound algorithm No extensive analyses to
Liu [15] Enhancing performance B: hypothetical biometric
imprints
M: Empirical, linear fusion
Highest authentication accuracy Vagueness in biometric
specification, not benchmarked
Venugopalan
[16]
Biometric
authentication
B: heart bear
M: Conceptual, classification
Supports critical authentications Induces computational complexity.
Jiping [17] Enhancing performance B: hypothetical biometric
imprints
M: Empirical, cryptography,
enhancing Das scheme [29]
Supports resisting higher range of
attacks
Doesn’t discusses the complexity
associated with key management
Frank [18] Biometric
authentication
B: Fingerprint
M: Analyzing touching behaviour
Performs continuous monitoring Accumulation and processing of too
much feature induces delay for long
run. Accuracy cannot be ensured for
multiple application usage pattern
Klonovs [19] Biometric
authentication
B: Brain signals
M: Experimental, feature-based
authentication
A novel mechanism Not applicable in practical as
accessibility depends upon stress
signal, which can happen to both
regular user & malicious user.
Simoens [20] Securing biometric
template
imprints
M: Conceptual, generic adversary
modeling
One framework supporting
multiple adversary mitigation
No numerical evidence to support
the claim.
Biggio [21] Impact of spoofing
attack
B: Facial, Fingerprint
M: Conceptual
Finding suggest higher
vulnerability of multimodal
compared to single modal, better
identification of adversary
No benchmarking
Urtiga [22] Biometric
authentication
B: Keystroke patterns
M: conceptual
Performs continuous monitoring Accumulation and processing of too
much feature induces delay for long
run. Accuracy cannot be ensured for
multiple application usage pattern
Sui and Du [23] Biometric
authentication
B: Iris
M: Empirical, feature extraction,
experimental
Support multiple level of security Accuracy depends on static images
only
Safie [24] Biometric
authentication
B: Heart beat
M: Empirical, feature extraction
Higher accuracy No extensive analyses to

Kumar [25] Biometric
authentication
B: Palm Print
M: Experimental, template
matching
Supports two biometric modalities No benchmark the outcome.
Chatterjee et al.
[26]
Biometric
authentication
B: posture information
M: Experimental, radial-basis
function
Offers higher accuracy No standardization or thresholding
for posture that can be mimicked
easily.
Veldhuis [27] Biometric
authentication
B: Face
M: Experimental,
Offers higher accuracy Success rate more for dataset and
less for real-time facial image.
Upmanyu [28] Biometric template
security
imprints
M: empirical, Blind
authentication, public key
encryption
Supports security over multiple
datasets
Communication overhead increases
for more rounds of query.
Apart from the above mentioned techniques, we also reviewed
other existing literatures as shown in Table 3 for exploring
different form of techniques and adopted modalities pertaining
to biometric based authentication. There are various
approaches that combinely uses different biometric to perform
authentication. From Fig. 3, it can be seen that many authors
have carried out the work on unimodal as well as multimodal
biometric system for strengthening authentication system.
These techniques were claimed to offer better security as well
as the better recognition rate of the system. However, it is still
an open end question to understand the best effectiveness in
the existing approaches.
TABLE 3 REVIEW OF MULTIMODAL BIOMETRIC SYSTEM.
Author Problem Technique Modalities
Soltanpour [30] To Improve face recognition and security. Local descriptors 2D-3D Face image
Oloyede [31] To improve the security
Stage in biometric system, architecture,
operation.
Uni and Multimodal
Haghighat [32] Reduce computational complexity. DCA, feature level fusion. Uni and Multimodal
Wan [33] To improve biometric recognition rate. DOA estimation, OMP, ADMM. PD Source diagnosis.
Martiri [34] To person authentication. Honey template and Bloom filters. Face image.
Sadhya [35] To secure database. Review paper Multimodal image
Bahrampour [36] Image classification
Multimodal task-driven dictionary learning
algorithm.
Multimodal image
Ding. [37] Image classification Comprehensive deep learning, NNs, SAE. Face
Almaadeed [38]
Authentication,
To identify speaker.
Neural network, wavelet transforms.,
MFCCs.
Voice
Yang [39]
To improve Recognition accuracy and
security level.
Mutual dependency Fingerprint and face.
DeCann[40] Re-identification
Mathematical Modelling errors in
Biometric.
Face and Finger print
Moutafis [41] To improve recognition rate. Rank based score normalization Face.
Meng [42] To protect user privet information Biometric identification on Mobile phones
Physiological and behavioral
characteristics.
Paul [43]
To improve quality, dimensionality
reduction and classifier.
Fisher Linear Discriminant Analysis Face, ear, signature
Zhang [44] Dimensionality reduction Semi supervised Learning Techniques. Face, finger print
Baig [45]
To improve speed and reduce memory
constraints.
Cascaded classifiers, Mahalanbis distance
algorithm.
Multimodal biometrics.
Murakami [46]
To improve Wolves and lambs accuracy
rate.
Optimal fusion algorithm. Face, Finger Print.
Shekhar[47] To improve recognition rate.
Multimodal sparse method and fusion based
methods.
Signature, Face.

Simon [48] To improve recognition rate, Deep learning convolution neural networks. Face, Signature.
Roy [49] identification accuracy Fuzzy C-means clustering iris, face
Canuto [50]
To overcome privacy concerns in
management.
Feature selection method. finger print, signature
Kumar [51] To improve Security BCH encoding and Hash function. Unimodal and Multimodal
Neverova [52] To Identity From Motion Patterns. Deep learning NN natural human kinematics
Poignant [53]
To identify unsupervised speakers, to
reduce cost.
Speaker diarization module. Speech, written names.
Mezai [54] To improve voice and face classification. PSO algorithm Face, Voice
From the above studies, it has been seen that facial image and
the fingerprint are the most common form of the biometric
attributes that has been considered for performing
authentication mechanism. Although, there are other attributes
e.g. iris, signature, speech, signature, etc are also attempted for
evolving up various authentication strategies, they were less in
numbers. It is believed that multimodal biometric is the next
generation method to strength the authentication potential of
unimodal biometrics and is likely to be adopted in faster pace
in near future.
IV. MODERN TECHNIQUES OF MULTI-MODAL
AUTHENTICATION
Usage of multimodal biometric is considered as one of the
potential method to ensure highly secured authentication
mechanism of biometrics with lesser chances of getting
compromised. The prime reason behind this is the usage of
more number of sophisticated biometric signals that are used
for performing encryption. We reviewed the trend of usage
patterns of using multiple forms of biometric signals where the
most frequently used signals are face, iris, ear image,
fingerprint, and signature.
Figure 3 Research Trend of Multimodal Biometrics
Fig.3 highlights the existing research trend where it can be
seen that usage of iris and face as well as usage of face and
fingerprint are most practiced methods in multimodal
biometric authentication. In this section, we give a brief of
usage methodology of some of the frequently exercised
multimodal biometrical authentication techniques as follows:
 Iris and Face: The internal structure of iris as well as
facial geometry is the easily available biometrical
attribute used for multimodal authentication. Usage
of such techniques has been witnessed in the work
carried out by Roy et al. [55], Ramachandran et al.
[56], Galbally et al. [57], and Tan et al. [58]. Fig.4
shows the common methodology found in the
techniques are extraction of features from iris as well
as face in order to check their similarity score. The
advantage of technique is that iris structure is hard to
be replicated or compromised and the disadvantage is
the facial geometry features will be required to be
consistently updated with the ages of changes of
facial structure. Hence, these cannot be used for long
run application and calls for constant update.
 Ear Image and Face: The combination of ear image
and facial structure offers a significant biometric
attribute owing to static structure of ear for any
human over a longer and longer period of time. The
work carried out by Javadtalab et al. [59] is the only
standard work that was published during 2010 till
date. Such techniques uses feature extraction from
both biometric image followed by feature fusion and
discriminates analysis (Fig.5). The advantage of this
technique is that biometric template pertaining to ear
of a specific subject doesn’t require any updates in its
template; however, it still suffers from frequent
updates of facial expression.
 Face, Fingerprint, and Voice: Apart from face and
voice, duplication of fingerprint is highly challenging
and hence this form of multimodal proves to offer
secured authentication mechanism. The work carried
out by Jain et al. [60] has used such technique. Fig.6
shows the methodology of such technique where a
template database is constructed for all three
biometric attribute followed by decision-based fusion
process to result in acceptance or rejection of query
of authentication. The advantage of such techniques
is its unique features that are very difficult to be
replicated at same time and its disadvantage is a
slight variation in pitch of voice can fail the
authentication system even for legitimate user.
 Face, Ear image, and Signature: Usage of face, ear
image, and signature is considered as one of the
simplistic form of biometric authentication owing to
its charecteristics of non-updation requirement in
regular interval. Fig.7 highlights the technique where
features of all the attributes are considered for
similarity check. Such work was carried out by

Monwar et al. [61] who have perform construction
analysis in social network using decision fusion.
None of the three attributes is expected to be changed
over longer period of time which poses a beneficial
factor while the demerit point is the possibility of
forged signature. Using three different photographs of
face, ear, and signature, even a malicious user can
have access to accounts. Hence, this process is not
safer.
 Face, Teeth Image, and Voice: Usage of facial image,
teeth image, and voice is another innovative
technique evolved up in recent times. The logic
behind such approach is to extract features using
different techniques on different biometric attributes
followed by score fusion techniques. The work
carried out by Kim et al. [62] is one such example
shown in Fig.8. The advantage of this technique is its
simpler usage and its disadvantage is related to its
sensitivity associated with the pitch of the voice.
 Fingerprint, Iris, and Face: This is one of the most
frequently used practices in multimodal biometric-
based authentication system. The work done by Ravi
et al.[63] and, Ko [64] has used similar approach. The
advantage of this mechanism is its speedy and precise
authentication mechanism and the disadvantage of
this mechanism is its consistent update of the
biometric dataset as it cannot identify forged
fingerprints and facial structure. A malicious user can
carry a photograph of a person and rubber print of
fingerprint to have an easy access to legitimate
person resource.
All the above mentioned techniques are quite a new beginning
of research attempts towards using multimodal biometric-
based authentication system. It is still in nascent stage of
development.
Figure 4 Usage of Iris & Facial Feature [55,56,57,58]
Figure 5 Usage of ear Image & Facial Feature [59]
Figure 6 Usage of Face, Fingerprint, and Voice [60]

Figure 7 Usage of Face, ear image, and Signature [61]
2D-DCT
2D-DCT
Pitch
MFCC
EHMM
Face
Database
EHMM
Teeth
Database
GMM
Voice
Database
Score
Normalize
Score
Fusion
Accept/
Reject
Figure 8 Usage of Face, teeth image, and voice [62]
Existing implementation of the all the above mentioned
methodologies towards multimodal biometric-based
authentication have been performed in standard datasets e.g.
FERET dataset [65], CASIA image dataset [66], CVRL
dataset [67], Biometric Recognition Group [68], Poly U Palm
print database, etc. These datasets have images pertaining to
face, fingerprint, palm, signature, iris, gait, etc. Out of all the
technique used in multi-modal forms, fingerprint is definitely
one of the dominant attribute being used in experiments. There
have been highest number studies towards the direction of
using fingerprints alone owing to the potential advantages
associated with it. Therefore, it can be said that fingerprint is
one of the essential biometric frequently considered in many
existing techniques of multimodal authentication system.
Combining fingerprint with other biometrics offered highest
level of security and faster authentication at same time.
Figure 9 Usage of Face, iris, and Fingerprint [63, 64]

V. OPEN RESEARCH CHALLENGES
Till prior section, various research-based techniques have been
discussed to show that there has been evolution of different
techniques and algorithm towards strengthening the biometric-
based authentication. This section discusses about the open
research challenges that were identified not to be addressed by
any existing system. Some of the existing challenges that are
unaddressed are as following:
 More Studies Towards Unimodal and Less Towards
Multimodal: Majority of the existing techniques of
biometrics have used single biometric attributes for
performing authentication. Although, some
researchers have also explored better feasibility of
using multi-modal approaches, but there are less
amount of research work in proportion for this. The
security complications used in unimodal is very
different than considered for multimodal that is also
found missing to be addressed in existing approaches
of multi-modal authentication.
 Lesser focus on Computational Complexity: Some of
the biometrics consumes enough data for both storing
and performing authentication. Hence, there is always
a complexity associated with processing algorithm in
order to perform similarity match. Although, many
system uses feature extracted for performing
similarity check, but the amount of feature (especially
in multi-modal authentication) is quite higher as
compared to unimodal authentication system.
However, almost none of the existing approaches
have ever considered computational complexity into
consideration while assessing the effectiveness of the
research outcomes.
 Less Emphasis to Iris: The potential usage of iris as
well as the internal networks of veins inside retina is
the best non-duplicable biometric attribute. However,
the adoption of iris is found in less proportion in
contrast to fingerprint, teeth image, ear image, voice,
etc. Moreover, usage of iris image has been always
considered from dataset and less work towards real-
time iris has been reported till date.
 More usage of Iterative Methods: Existing
approaches are mainly found to follow highly
iterative mechanism for performing algorithm
implementation. Usage of iterative has good effect on
accuracy but adverse effect on system performance
e.g. response time, energy consumption, resource
dependencies, memory. Hence, iterative methods are
not much found to be enhanced in order to solve such
system performance problems in any of the existing
literatures.
VI. CONCLUSION
The proposed paper mainly aims to study the effectiveness
of the existing approaches towards enhancing the biometric-
based authentication system. The study contributions of the
presented paper are as follows: i) Existing review-based studies
toward effectiveness of existing biometric-based authentication
mechanism is very less to find. Available review works
towards biometric doesn’t discuss about limitations and was
not found to emphasize on multimodal approaches. Therefore,
the first contribution of proposed study is that this paper offers
a wide visualization of the study effectiveness of complete
biometric based authentication system be it unimodal or
multimodal system. ii) The second study contribution is the
essential findings of proposed investigation e.g. open issues in
biometric design principle, usage of symptomatic approaches,
multimodal approaches still being underrated, less extent of
studies towards multimodal as compared to unimodal based
authentication, etc. Therefore, our future work will be to
address some of the significant issues. We will initiate our
investigations towards developing a novel framework that
offers highly resilient authentication using iris as there has been
no significant enhancement in adopting iris in existing system.
An attempt will be to evolve with a novel combinatorial-based
optimization approach where the data integrity will be
emphasized along with biometric template security.
REFERENCES
[1]. Virginio Cantoni, Dimo Dimov, Massimo Tistarelli, “Biometric
Authentication: First International Workshop, BIOMET 2014,
Sofia, Bulgaria, June 23-24, 2014. Revised Selected Papers”,
Springer, pp. 265, 2014
[2]. Forbes Technology Council, “The Promise And Challenges Of
Biometrics”, Forbes Community Voice, 2016
[3]. Aumi, Md Tanvir Islam, and Sven Kratz. "AirAuth: towards
attack-resilient biometric authentication using in-air gestures."
In Proceedings of the extended abstracts of the 32nd annual ACM
conference on Human factors in computing systems, pp. 1585-
1590. ACM, 2014.
[4]. Pflug, Anika, and Christoph Busch. "Ear biometrics: a survey of
detection, feature extraction and recognition methods." IET
biometrics 1, no. 2 (2012): 114-129.
[5]. Arun A. Ross, Karthik Nandakumar and Anil K. Jain, "Handbook
of Multibiometrics (international Series On Biometrics)", Springer
science and Business Media, YES edition, 2011.
[6]. B. Arslan, E. Yorulmaz, B. Akca and S. Sagiroglu, "Security
Perspective of Biometric Recognition and Machine Learning
Techniques," 2016 15th IEEE International Conference on
Machine Learning and Applications (ICMLA), Anaheim, CA,
2016, pp. 492-497.
[7]. Pandey, Shubham, Ruchi Varshney, Shivani Gupta, Neha
Chaudhary, and Sangeeta Singh. "Study of a Secure
Communication System." Imperial Journal of Interdisciplinary
Research 2, no. 11 (2016).
[8]. Mellado, Daniel, ed. IT Security Governance Innovations: Theory
and Research: Theory and Research. IGI Global, 2012.
[9]. S. Ghosh, A. Majumder, J. Goswami, A. Kumar, S. P. Mohanty
and B. K. Bhattacharyya, "Swing-Pay: One Card Meets All User
Payment and Identity Needs: A Digital Card Module using NFC
and Biometric Authentication for Peer-to-Peer Payment," in IEEE
Consumer Electronics Magazine, vol. 6, no. 1, pp. 82-93, Jan.
2017.
[10]. Y. Choi, Y. Lee, and D. Won, "Security improvement on biometric
based authentication scheme for wireless sensor networks using
fuzzy extraction", International Journal of Distributed Sensor
Networks, 2016
[11]. H. S. Choi, B. Lee and S. Yoon, "Biometric Authentication Using
Noisy Electrocardiograms Acquired by Mobile Sensors," in IEEE
Access, vol. 4, no. , pp. 1266-1273, 2016.

[12]. Bhuiyan, A. Hussain, A. Mian, T. Y. Wong, K. Ramamohanarao
and Y. Kanagasingam, "Biometric authentication system using
retinal vessel pattern and geometric hashing," in IET Biometrics,
vol. 6, no. 2, pp. 79-88, 3 2017.
[13]. J. Šeděnka, S. Govindarajan, P. Gasti and K. S. Balagani, "Secure
Outsourced Biometric Authentication With Performance
Evaluation on Smartphones," in IEEE Transactions on Information
Forensics and Security, vol. 10, no. 2, pp. 384-396, Feb. 2015.
[14]. W. Kang, D. Cao and N. Liu, "Deception With Side Information in
Biometric Authentication Systems," in IEEE Transactions on
Information Theory, vol. 61, no. 3, pp. 1344-1350, March 2015.
[15]. Y. Liu, Z. Yang, C. Y. Suen and L. Yang, "A Study on
Performance Improvement Due to Linear Fusion in Biometric
Authentication Tasks," in IEEE Transactions on Systems, Man,
and Cybernetics: Systems, vol. 46, no. 9, pp. 1252-1264, Sept.
2016.
[16]. S. Venugopalan, M. Savvides, M. O. Griofa and K. Cohen,
"Analysis of Low-Dimensional Radio-Frequency Impedance-
Based Cardio-Synchronous Waveforms for Biometric
Authentication," in IEEE Transactions on Biomedical Engineering,
vol. 61, no. 8, pp. 2324-2335, Aug. 2014.
[17]. L. Jiping, D. Yaoming, X. Zenggang, and L. Shouyin, "An
improved biometric-based user authentication scheme for C/S
system", International Journal of Distributed Sensor Networks,
2014
[18]. M. Frank, R. Biedert, E. Ma, I. Martinovic and D. Song,
"Touchalytics: On the Applicability of Touchscreen Input as a
Behavioral Biometric for Continuous Authentication," in IEEE
Transactions on Information Forensics and Security, vol. 8, no. 1,
pp. 136-148, Jan. 2013.
[19]. J. Klonovs, C. K. Petersen, H. Olesen and A. Hammershoj, "ID
Proof on the Go: Development of a Mobile EEG-Based Biometric
Authentication System," in IEEE Vehicular Technology Magazine,
vol. 8, no. 1, pp. 81-89, March 2013.
[20]. K. Simoens, J. Bringer, H. Chabanne and S. Seys, "A Framework
for Analyzing Template Security and Privacy in Biometric
Authentication Systems," in IEEE Transactions on Information
Forensics and Security, vol. 7, no. 2, pp. 833-841, April 2012.
[21]. Biggio, Z. Akhtar, G. Fumera, G. L. Marcialis and F. Roli,
"Security evaluation of biometric authentication systems under real
spoofing attacks," in IET Biometrics, vol. 1, no. 1, pp. 11-24,
March 2012.
[22]. E. V. Cunha Urtiga and E. D. Moreno, "Keystroke-based biometric
authentication in mobile devices," in IEEE Latin America
Transactions, vol. 9, no. 3, pp. 368-375, June 2011.
[23]. Y. Sui, X. Zou and E. Y. Du, "Biometrics-Based Authentication: A
New Approach," 2011 Proceedings of 20th International
Conference on Computer Communications and Networks
(ICCCN), Maui, HI, 2011, pp. 1-6.
[24]. S. I. Safie, J. J. Soraghan and L. Petropoulakis, "Electrocardiogram
(ECG) Biometric Authentication Using Pulse Active Ratio (PAR),"
in IEEE Transactions on Information Forensics and Security, vol.
6, no. 4, pp. 1315-1322, Dec. 2011.
[25]. Kumar and D. Zhang, "Improving Biometric Authentication
Performance From the User Quality," in IEEE Transactions on
Instrumentation and Measurement, vol. 59, no. 3, pp. 730-735,
March 2010.
[26]. Chatterjee, R. Fournier, A. Nait-Ali and P. Siarry, "A Postural
Information-Based Biometric Authentication System Employing S-
Transform, Radial Basis Function Network, and Extended Kalman
Filtering," in IEEE Transactions on Instrumentation and
Measurement, vol. 59, no. 12, pp. 3131-3138, Dec. 2010.
[27]. Q. Tao and R. Veldhuis, "Biometric Authentication System on
Mobile Personal Devices," in IEEE Transactions on
Instrumentation and Measurement, vol. 59, no. 4, pp. 763-773,
April 2010.
[28]. M. Upmanyu, A. M. Namboodiri, K. Srinathan and C. V. Jawahar,
"Blind Authentication: A Secure Crypto-Biometric Verification
Protocol," in IEEE Transactions on Information Forensics and
Security, vol. 5, no. 2, pp. 255-268, June 2010.
[29]. Ross and A. Abaza, "Human Ear Recognition," in Computer, vol.
44, no. 11, pp. 79-81, Nov. 2011.
[30]. S. Soltanpour and Q. J. Wu, "Multimodal 2D–3D face recognition
using local descriptors: pyramidal shape map and structural
context," in IET Biometrics, vol. 6, no. 1, pp. 27-35, 1 2017.
[31]. M. O. Oloyede and G. P. Hancke, "Unimodal and Multimodal
Biometric Sensing Systems: A Review," in IEEE Access, vol. 4,
no. , pp. 7532-7555, 2016
[32]. M. Haghighat, M. Abdel-Mottaleb and W. Alhalabi, "Discriminant
Correlation Analysis: Real-Time Feature Level Fusion for
Multimodal Biometric Recognition," in IEEE Transactions on
Information Forensics and Security, vol. 11, no. 9, pp. 1984-1996,
Sept. 2016.
[33]. L. Wan, G. Han, L. Shu, S. Chan and N. Feng, "PD Source
Diagnosis and Localization in Industrial High-Voltage Insulation
System via Multimodal Joint Sparse Representation," in IEEE
Transactions on Industrial Electronics, vol. 63, no. 4, pp. 2506-
2516, April 2016.
[34]. E. Martiri, M. Gomez-Barrero, B. Yang and C. Busch, "Biometric
template protection based on Bloom filters and honey templates,"
in IET Biometrics, vol. 6, no. 1, pp. 19-26, 1 2017.
[35]. Sadhya, S. K. Singh and B. Chakraborty, "Review of key-binding-
based biometric data protection schemes," in IET Biometrics, vol.
5, no. 4, pp. 263-275, 12 2016.
[36]. S. Bahrampour, N. M. Nasrabadi, A. Ray and W. K. Jenkins,
"Multimodal Task-Driven Dictionary Learning for Image
Classification," in IEEE Transactions on Image Processing, vol.
25, no. 1, pp. 24-38, Jan. 2016.
[37]. Ding and D. Tao, "Robust Face Recognition via Multimodal Deep
Face Representation," in IEEE Transactions on Multimedia, vol.
17, no. 11, pp. 2049-2058, Nov. 2015.
[38]. N. Almaadeed, A. Aggoun and A. Amira, "Speaker identification
using multimodal neural networks and wavelet analysis," in IET
Biometrics, vol. 4, no. 1, pp. 18-28, 3 2015.
[39]. W. Yang, J. Hu, S. Wang and C. Chen, "Mutual dependency of
features in multimodal biometric systems," in Electronics Letters,
vol. 51, no. 3, pp. 234-235, 2 5 2015.
[40]. DeCann and A. Ross, "Modelling errors in a biometric re-
identification system," in IET Biometrics, vol. 4, no. 4, pp. 209-
219, 12 2015.
[41]. P. Moutafis and I. A. Kakadiaris, "Can We Do Better in Unimodal
Biometric Systems? A Rank-Based Score Normalization
Framework," in IEEE Transactions on Cybernetics, vol. 45, no. 12,
pp. 2654-2667, Dec. 2015.
[42]. W. Meng, D. S. Wong, S. Furnell and J. Zhou, "Surveying the
Development of Biometric User Authentication on Mobile
Phones," in IEEE Communications Surveys & Tutorials, vol. 17,
no. 3, pp. 1268-1293, thirdquarter 2015.
[43]. P. P. Paul, M. L. Gavrilova and R. Alhajj, "Decision Fusion for
Multimodal Biometrics Using Social Network Analysis," in IEEE
Transactions on Systems, Man, and Cybernetics: Systems, vol. 44,
no. 11, pp. 1522-1533, Nov. 2014.
[44]. Q. Zhang, Y. Yin, D. C. Zhan and J. Peng, "A Novel Serial
Multimodal Biometrics Framework Based on Semisupervised
Learning Techniques," in IEEE Transactions on Information
Forensics and Security, vol. 9, no. 10, pp. 1681-1694, Oct. 2014.
[45]. Baig, A. Bouridane, F. Kurugollu and B. Albesher, "Cascaded
multimodal biometric recognition framework," in IET Biometrics,
vol. 3, no. 1, pp. 16-28, March 2014.
[46]. T. Murakami, K. Takahashi and K. Matsuura, "Toward Optimal
Fusion Algorithms With Security Against Wolves and Lambs in
Biometrics," in IEEE Transactions on Information Forensics and
Security, vol. 9, no. 2, pp. 259-271, Feb. 2014.
[47]. S. Shekhar, V. M. Patel, N. M. Nasrabadi and R. Chellappa, "Joint
Sparse Representation for Robust Multimodal Biometrics

Recognition," in IEEE Transactions on Pattern Analysis and
Machine Intelligence, vol. 36, no. 1, pp. 113-126, Jan. 2014.
[48]. M. O. Simón et al., "Improved RGB-D-T based face recognition,"
in IET Biometrics, vol. 5, no. 4, pp. 297-303, 12 2016.
[49]. K. Roy, J. Shelton, B. O'Connor and M. S. Kamel, "Multibiometric
system using fuzzy level set, and genetic and evolutionary feature
extraction," in IET Biometrics, vol. 4, no. 3, pp. 151-161, 9 2015.
[50]. M. de Paula Canuto, M. C. Fairhurst and F. Pintro, "Ensemble
systems and cancellable transformations for multibiometric-based
identification," in IET Biometrics, vol. 3, no. 1, pp. 29-40, March
2014.
[51]. Kumar and A. Kumar, "A Cell-Array-Based Multibiometric
Cryptosystem," in IEEE Access, vol. 4, no. , pp. 15-25, 2016.
[52]. N. Neverova et al., "Learning Human Identity From Motion
Patterns," in IEEE Access, vol. 4, no. , pp. 1810-1820, 2016.
[53]. J. Poignant, L. Besacier and G. Quénot, "Unsupervised Speaker
Identification in TV Broadcast Based on Written Names," in
IEEE/ACM Transactions on Audio, Speech, and Language
Processing, vol. 23, no. 1, pp. 57-68, Jan. 2015.
[54]. L. Mezai and F. Hachouf, "Score-Level Fusion of Face and Voice
Using Particle Swarm Optimization and Belief Functions," in IEEE
Transactions on Human-Machine Systems, vol. 45, no. 6, pp. 761-
772, Dec. 2015.
[55]. K. Roy, J. Shelton, B. O'Connor and M. S. Kamel, "Multibiometric
system using fuzzy level set, and genetic and evolutionary feature
extraction," in IET Biometrics, vol. 4, no. 3, pp. 151-161, 9 2015.
[56]. R. P. Ramachandran et al., "Vertical Integration of Biometrics
Across the Curriculum: Case Study of Speaker, Face and Iris
Recognition," in IEEE Circuits and Systems Magazine, vol. 14, no.
3, pp. 55-69, thirdquarter 2014.
[57]. J. Galbally, S. Marcel and J. Fierrez, "Image Quality Assessment
for Fake Biometric Detection: Application to Iris, Fingerprint, and
Face Recognition," in IEEE Transactions on Image Processing,
vol. 23, no. 2, pp. 710-724, Feb. 2014.
[58]. W. Tan and A. Kumar, "Efficient and Accurate At-a-Distance Iris
Recognition Using Geometric Key-Based Iris Encoding," in IEEE
Transactions on Information Forensics and Security, vol. 9, no. 9,
pp. 1518-1526, Sept. 2014.
[59]. S. Tiwari, S. Jain, S. S. Chandel, S. Kumar and S. Kumar,
"Comparison of adult and newborn ear images for biometric
recognition," 2016 Fourth International Conference on Parallel,
Distributed and Grid Computing (PDGC), Waknaghat, India,
2016, pp. 421-426.
[60]. K. Jain, "Biometrics: Proving Ground for Image and Pattern
Recognition," Fourth International Conference on Image and
Graphics (ICIG 2007), Sichuan, 2007, pp. 3-3.
[61]. M. M. Monwar and M. L. Gavrilova, "Multimodal Biometric
System Using Rank-Level Fusion Approach," in IEEE
Transactions on Systems, Man, and Cybernetics, Part B
(Cybernetics), vol. 39, no. 4, pp. 867-878, Aug. 2009.
[62]. J. Kim, K. W. Chung and K. S. Hong, "Person authentication using
face, teeth and voice modalities for mobile device security,"
in IEEE Transactions on Consumer Electronics, vol. 56, no. 4, pp.
2678-2685, November 2010.
[63]. S. Ravi and D. P. Mankame, "Multimodal biometric approach
using fingerprint, face and enhanced iris features
recognition," 2013 International Conference on Circuits, Power
and Computing Technologies (ICCPCT), Nagercoil, 2013, pp.
1143-1150.
[64]. Ko, Teddy. "Multimodal biometric identification for large user
population using fingerprint, face and iris recognition." In Applied
Imagery and Pattern Recognition Workshop, 2005. Proceedings.
34th, pp. 6-pp. IEEE, 2005.
[65]. Phillips PJ, Moon H, Rizvi SA, Rauss PJ. The FERET evaluation
methodology for face-recognition algorithms. IEEE Transactions
on pattern analysis and machine intelligence. 2000
Oct;22(10):1090-104.
[66]. Proenca, H., Filipe, S., Santos, R., Oliveira, J. and Alexandre, L.A.,
2010. The ubiris. v2: A database of visible wavelength iris images
captured on-the-move and at-a-distance. IEEE Transactions on
Pattern Analysis and Machine Intelligence, 32(8), p.1529.
[67]. G. Santos, P. T. Fiadeiro and H. Proença, "BioHDD: a dataset for
studying biometric identification on heavily degraded data," in IET
Biometrics, vol. 4, no. 1, pp. 1-9, 3 2015.
[68]. Phillips, P. Jonathon, Patrick J. Flynn, and Kevin W. Bowyer.
"Lessons from collecting a million biometric samples." Image and
Vision Computing 58 (2017): 96-107.

An Investigation towards Effectiveness of Present State of Biometric-Based Authentication System

More Related Content

What's hot (19)

Similar to An Investigation towards Effectiveness of Present State of Biometric-Based Authentication System (20)

More from RSIS International (20)

Recently uploaded (20)

An Investigation towards Effectiveness of Present State of Biometric-Based Authentication System