CVPR Daily - Wednesday

When did you start teaching at George Mason University? Less than a year ago, August 2017. My tenure track just started. Can you tell us about your work? I teach cybersecurity, and I will teach machine learning in the fall. Mainly, I do research. I am a research professor. My expertise is in biometrics so I am writing proposals on that. Then I also collaborate with my colleagues on cybersecurity projects. Do you want to share something about biometrics that we don’t know? Oy! Ah, well, I don’t know if you don’t know, but I can share what I’m doing. Please do! I graduated in 2010, and I worked on multi-biometric systems and their security. Then I started my postdoc so I moved over to the US. I worked with Arun Ross at West Virginia University and with Bojan Cukik later on. Now I am on my own with my students. You did a long postdoc, right? Yes, for five years. I did enjoy it. Now, my interests are on biases that affect security performance on biometric systems. The modality I’m working on is fingerprint. There are three aspects that I’m very interested in. One is the capture bias so I work on cross sensor matching to enhance the performance. Then I also work on biases on demographics, which are gender, age, and ethnicity. I also work on biases for security. There is, of course, security issues related to biometric systems because they are subject to spoof attacks. They replicate the biometric trait and present it to the system so that they can actually deceive it and gain unauthorized access. We know that in applications like at the airport or any access control, this is very dangerous. This is very, very dangerous because biometrics are employed to identify suspected terrorists. Even last year, I think, the number of civilians killed in terrorist attacks was still over 18,000. So we need to work on these biases. What body parts are you working on in biometrics? The iris? Right now, I’m working on fingerprints, Emanuela Marasco Emanuela Marasco is Assistant Professor at George Mason University in the Department of Information Science and Technology. 12 Women in Science Wednesday

RkJQdWJsaXNoZXIy NTc3NzU=