Research
Current Activities
My current research focus lies within the
fields of information security and privacy enhancing technology.
I'm particularly interested in probing areas where it seems that,
rather than helping to address existing security and privacy problems,
developments in information technology have tended to exacerbate
these problems, or have created entirely new ones. (See my
links
page to find out more on these kinds of issues, beyond my own
research.)
My current research
is exploring these issues in the context of biometrics.
Biometrics
My dissertation research
focuses on developing countermeasures to biometric technologies. In
designing secure systems, it is just as important to understand the
limitations of a given component as it is to understand its capabilities.
If you do not know how part of your secure system is vulnerable, you can
be sure that your adversary, i.e. those parties that your secure system
is trying to keep out, will be highly motivated to identify those
vulnerabilities for you. Identifying weaknesses proactively is the
best way to learn how, and how not, to make use biometric identification
technology in building a secure system.
It is also a good idea to guard against the possibility of our technology
being used against us.
Misuse of biometric technology is certainly a serious concern for
personal privacy,
but consider also the possible impacts of biometrics on, for example,
witness protection, or
the activities of covert agents. It is easy to envision scenarios in these
domains where involuntary identification could be very dangerous, even
life-threatening.
Abuse of biometrics is particularly worrisome with so-called
human-id at a distance technologies,
which can allow identification of individuals without their knowledge or
consent. The simplicity and always-falling cost of storing digital recordings
of video, still images, and other identifying information
means that the surreptitious identification need not even
happen immediately. On the contrary: identification can happen years
after an image is recorded, while biometric technologies continue to improve
in the meantime.
In order to begin addressing this problem, I am working on developing
techniques to remove, conceal, or otherwise degrade biometric information.
Additionally, I am developing a metric suitable
for a principled cost-benefit analysis of any biometric obfuscation technique,
and validating my results empirically.
Projects
I am currently running the Masks Project. This study aims to improve privacy and security of face
recognition technology by identifying which parts of the human face are
most distinctive, and learning what is required to hide some of this information
should the situation require it. You can find out more about what is happening with the project
here.
Papers
- James Alexander and Jonathan Smith, Engineering Privacy in Public: Confounding Face Recognition, In The Third Workshop on Privacy
Enhancing Technologies, 2003.
- James Alexander, The MASKS Face Image Data Set: Collection Procedures and Progress Report, Technical Report MS-CIS-07-03, University of Pennsylvania, 2007.
-
James Michael Alexander, MASKS: Maintaining Anonymity by Sequestering Key Statistics, Publicly accessible Penn Dissertations, Paper 32, http://repository.upenn.edu/edissertations/32/, 2009.
Back to home page
My public key
jalex@cis.upenn.edu
Last Modified:
November 15, 2009 (04:20:24 PM).