Vous êtes sur la page 1sur 2

Initial Statement of Research Intent

Notes to the applicant:

 If your application progresses you may be asked to submit a more detailed Statement of
Research Intent.
 Applicants should note that students are expected to submit their thesis after 3-4 years full-
time study (6-8 years part-time).

Your details

Annisa Istiqomah Arrahmah

Name of applicant:

Student ID (if known):

Department/s (if known):

Electrical & Computer Engineering

Associate Professor Waleed H. Abdulla

Proposed supervisor/s (if known):

Area of research interest

Provisional title of thesis or area of research interest, and a brief description of the proposed

I have a deep interest on how the technologies can be used to improve the quality of human health
care, unravel the secrets of our body, and uniquely identify each person based on the unique
characteristics embedded in their body. As an electrical engineering undergraduate student, I had
a chance to learn more about image processing where I did my undergraduate thesis to enhance
the image quality input of human health care devices for the next processing module. It was a
challenging, yet exciting project which led me to explore more about human body parts and its
uniqueness. I found out that several parts of human body have unique properties that can be
extracted for identification purposes using image processing.

After finishing my bachelor's degree, I took a master's degree in security and information systems
field. During my study, I learned that information system involves data about people. In order to
associate the data with a certain human, a reliable, effective and efficient identification scheme is
needed to be established and maintained. There are several techniques in human identification
including names, codes, token and knowledge-based identification, also biometrics. Based on my
previous experience and my interest while I was in my undergraduate study, I chose biometrics
as personal recognition in information systems as my main thesis topic. By using biometrics, an
automatic and convenience personal recognition can be established. However, because this
method involves human physiological or behavioral characteristics, it is also vulnerable for being
forged. Once the biometric features are compromised, the user will not have any substitute for it.
Based on this problem, my thesis concern was about the security part of the biometric recognition
system. The research purpose was to enhance security in a key binding process by using biometric
cryptography. By using this method, the user’s biometric template is protected while the matching
process was established. However, this method can make the recognition process taking more
time for higher accuracy.

Improving security aspects in biometric recognition systems not only by adding more security
modules and assessing it but also by considering the attacker's behavior from the adversaries’
side. Research can be done by inspecting the vulnerabilities of the systems along with the
corresponding attack, then find the countermeasures and make the defense mechanisms.
Nowadays, innovations in machine learning have influenced every part of research development.
The increasing of machine learning techniques also gives impact in research about biometric
recognition systems. By using machine learning, attacks against biometric recognition systems
can be learned by creating adaptive and intelligent adversaries to gain access to the biometric
systems. Based on this phenomenon, my next research intent is to explore about biometric
recognition system using adversarial machine learning method. First, a specific algorithm for

Last updated: 17 April 2013 Page 1 of 2

Initial Statement of Research Intent
biometric recognition is chosen as a resource. By identifying specific attack that could penetrate
the biometric recognition system, an adaptive adversaries will be made based on it. The specific
attack could be a spoofing attack, denial of service or any other sophisticated attacks of the
biometric algorithms used in the systems. After that, a corresponding countermeasures will be
analyzed to cover the vulnerabilities. These countermeasures are already proposed in the field of
adversarial machine learning.

In Indonesia, biometric recognition has been used in various systems although not in the crucial
ones such as financial transactions due to regulation problems. Nevertheless, in the future it is
possible that biometric recognition will dominate the identification process so that the development
of the biometric recognition system still needed to make a robust system. Unfortunately, research
in this field in Indonesia still rare and there is no specific research group in this field. After I
searched, I found that in Auckland, there is a professor that has research focused on biometric.
By entering the research group at University of Auckland, I hope that I can get an adequate
research environment and research colleagues whom I can share my thought and get guidance
from the experts there.

Research/professional background

Briefly summarise any previous research/professional experience in your area of interest

I am a graduated double-degree master student from Information Security, Institut Teknologi

Bandung, Indonesia (graduated in March 2017) and Interdisciplinary Program of Information
System (graduated in February 2017), Pukyong National University, South Korea. My thesis was
about fingerprint authentication using fuzzy vault biometric cryptosystem. The fuzzy vault
biometric cryptosystem is used to enhance the security aspects in fingerprint authentication
system. In my thesis, I modified the fuzzy vault non-random chaff generator for fingerprint to
enhance the security performance in authentication process. The fuzzy vault non-random chaff
generator was simulated in Matlab. After graduated, I became an academic assistant in Electrical
Engineering Department ITB. I also joined the research team there. My research project was about
Side-Channel Analysis in Embedded System. The Side Channel Analysis is a process to attack
cryptography algorithm. In my research, we attacked the AES encryption implemented in ATMega
128P. Previously in undergraduate study, my final project was about image processing in text to
speech devices for patient with low vision. I used Raspberry Pi as the platform and python
language for the programming language. Based on my previous experience, I believe that it will
help me accomplish my proposed research.

Last updated: 17 April 2013 Page 2 of 2