Lael Rayfield, Mathematics
“Simple Analysis of Biometric Matchers Based on Generated CMC Curves”
What are biometric matchers? In Forensic Science, biometrics refers to technologies that measure and analyze human body characteristics, such as DNA, fingerprints, eye retinas and irises, voice patterns, and facial patterns, for authentication purposes. Biometric matchers compare a person’s features against stored biometric features in a database to see if the computer can find a match and identify the subject. Modeling the performance of the automated matchers is important for recognizing abnormal behavior in test results, and predicting performance of matchers when the size of the database is greatly increased. In this talk I will present the results of comparing two different functions that model the performance of biometric matchers: 1) experimentally measured Cumulative Match Characteristic (CMC) curves; 2) and best-fit curves produced by the Power-law Distribution.
How did you find out that you could do research in your field in the summer?
I found out that there were research opportunities available for mathematics majors through the Meyerhoff Scholarship Program.
How did you know that research at the National Institute of Standards and Technology (NIST) was what you wanted to do?
I was interested in NIST because I thought it was a great opportunity to work alongside government researchers on a real problem that has not been solved. I also heard a lot of positive feedback from students who had participated in the SURF program.
Did you apply to other places?
Yes, I did! I applied to 16 other programs in addition to NIST and received three acceptances. I was also accepted into the Summer Institute for Training in Biostatistics at University of South Florida and the Summer Undergraduate Math Research program at Kansas Sate University. It is difficult to get a research position as a freshman, so I was very grateful to have options.
Was the application difficult to do? Did you have help with this?
I started my applications pretty early—by Thanksgiving I had a list of all 17 places I wanted to apply to and had written my personal statement. I started filling out the applications in early January. Some of the applications were more time consuming or required more written essays than others. I worked very diligently, though, and finished them all in about two weeks. I did have assistance from the Meyerhoff staff and my peer advisors in selecting good programs and writing a strong personal statement. I imagine the process would have been much more frustrating without their help!
What was your summer research project?
My research was on Biometric matchers. Biometric matchers compare images of a person’s features (such as DNA, fingerprints or irises) against images stored in a database to see if the computer can find a match and identify the subject. The matchers I analyzed were used for latent fingerprints and face recognition. I compared two different functions that model the performance of biometric matchers and tried to figure out which matchers worked best and why.
Who was your mentor for this project?
My mentor was Dr. Vladimir Dvornycheko, one of the mathematicians working on this project in the Information Access Division at NIST.
How much time did you put into this work?
NIST employees typically work 40 hours per week. I work from 8:30 to 5:00 on weekdays for 11 weeks this summer.
Were you paid? Where did you live?
Yes, I was paid! I’ve never had a real job before, so the excitement of receiving a paycheck was a new experience for me. All SURF students are paid a stipend. As for living arrangements, I lived in the Hyatt House hotel in Gaithersburg for free. It is nice to live so close to NIST and to meet aspiring researchers from all over the world.
What academic background did you have before you started?
I had just finished my freshman year when I started at NIST. I did not know very much about math research or how it is done until this summer.
How did you learn what you needed to know for this project?
My mentor and officemates were very patient with me and helped me get up to speed. I also spent some time reading books and research papers related to my project. Within a few weeks, I had a pretty solid understanding of everything.
What was the hardest part about your research?
The hardest part was probably the beginning. Initially I found myself staring at dozens of spreadsheets full of data, not having a clue what to do. I did make some mistakes, but over time I got used to working with the data.
What was the most unexpected thing?
I didn’t expect people at NIST to be so friendly. Everyone you meet to is happy to help you and very enthusiastic about what they do.
What is your advice to other students about getting involved in research?
Don’t be afraid to apply for a research position! Even if you don’t have much experience (like me), it doesn’t hurt to just apply. Also, don’t limit yourself by applying to only a few schools. Programs are competitive and it’s good to have as many options as possible.
What are your career goals?
After I finish at UMBC, I plan to go to graduate school for a Ph.D. I’m not sure what kind of career I’m suited for yet, but I have a lot of options to choose from. I’m interested in a career that involves math or math research.