Meet Jacob Biloki
Get to know our Biometrics Team - Spring 2017
Behind the scenes
On a sunny Wednesday afternoon, as I arrived to the Cybersecurity Center, I was greeted by some students working in diverse mobile biometric projects. Promptly, when I asked the students for Jacob, he looks at me and says: “Hi I am Jacob, I was waiting for you.” Not only was Jacob on time but he also greeted me formally with a confident smile. This was the first time I saw Jacob in person.
As I prepared my equipment for the interview, I observed Jacob interacting with his peers. He was sitting on a conference table looking at his computer and talking to the others students. He was always with that smile in his face that tells it all: He has passion for his work and he is brilliant!
INTRODUCTION
ECS Cyber: Jacob thank you for having us today in the Cybersecurity Center for your interview. It is a pleasure to be in a little piece of your world, thank you for allowing us to see it. I am very excited to get your story out and let people to know you better. Tell us a little bit of yourself, your major and plans. What is your passion?
Jacob: I am Jacob Biloki, a Computer Scientist in the concentration of machine learning and software engineering. I graduate December 2017. I plan to hopefully proceed to my masters in machine learning/Datamining and pattern recognition. I have been at CSUF for four years.
ABOUT THE MOBILE BIOMETRICS' RESEARCH
ECS Cyber: What is the impact of your mobile biometrics study in our society? How is it important?
Jacob: The impact of our current research of mobile biometrics is important due to the insecurities that come with unimodal biometrics. A unimodal approach is not nearly secure enough to protect the user’s data. There have been many methods designed to spoof all different types of unimodal systems from finger prints (the gummy bear hack), to face recognition being able to use a picture of the correct person. By breaking into the word of multimodal biometrics we want to help break out of this and ensure that users can trust a biometric system to secure their data. We can use more than one biometric such as face/voice to cross validate the user therefore making it more difficult to spoof.
ABOUT BUILDING THE BIOMETRICS MOBILE APP FOR A SAMSUNG GALAXY S7
ECS Cyber: How did you build the biometrics mobile app?
Jacob: The data collection app was a week-long process of feedback from the overseeing professor Dr. Gofman and ensuring that I have implemented the features requested in the application. I begin most of my mobile applications by building the back end, in this case how the user would take the picture and save it into an SQLite database. From there the user interface can be simple as this application is only for data collection. The final application of implementation of multimodal biometrics is another obstacle I am preparing to overcome, I have begun thinking of approaches to make the usability more fluid and easy to understand for the end user. And then comes the task of implementing computationally heavy machine learning methods on a mobile device with a limited battery life. This is something I am still contemplating and trying to find the best approach.
ECS Cyber: Who would be using it?
Jacob: I would for sure be using it, but hopefully everyone that cares about their data.
ECS Cyber: What programs or tools are you using in your research?
Jacob: For android development I am using Android Studio with the built in SQLite as the database. For the machine learning applications our testing is done on MATLAB.
ECS Cyber: What are the challenges you have or are facing with your biometrics project?
Jacob: Implementing the computationally heavy machine learning methods onto a limited device.
ECS Cyber: When will be project be finished?
Jacob: It is hard to say but we have made great strides in the past year.
JACOB'S PLANS FOR THE FUTURE
ECS Cyber: What are your plans when you graduate?
Jacob: I hope to find a good job and continue my research in the area of machine learning. The job I get has not been clear to me yet. Without a master’s degree in machine learning or at least five years of experience I doubt that it is a reliable career path. I look to possibly start as a software engineer.
JACOB'S FAVORITES
ECS Cyber: Android or Apple? Windows, Mac or Linux?
Jacob: Android and Linux for sure!
ECS Cyber: Something you live by, an anedote, or your favorite quote.
Jacob: "It’s not that I’m so smart, it’s just that I stay with problems longer. "
-Albert Einstein
JACOB'S HOBBIES
ECS Cyber: What are your hobbies?
Jacob: My hobbies are all in the realm of computer science, I occasionally reward myself after exams with playing video games such as league of legends but mostly I like to do my own research. I have recently pushed a new application to GitHub documenting my work on android devices using machine learning. I currently have implemented a MLP (multilayer perceptron) trained by the Wisconsin Breast Cancer data set hosted by UCI. To support the short battery life and power of a mobile device the learning methods are done on my server. The phone is an interface to send a request to the server, which will get fed forward though the trained algorithm. I hope to expand on this idea and continue my research. You can find my work here:
https://github.com/Jbiloki/MLServe
I am open to any criticism and ideas. I hope to further implement new machine learning methods, even on the currently used database. I believe that I can be more accurate with a model such as SVM to help further separate the classes. Thank you.
ECS Cyber: It was a pleasure to meet you in person and get to know you a little more. Thank you Jacob for giving us the opportunity to interview you!
Jacob: Thank you for taking the time and interest to talk to me I enjoyed this interview and hope we can continue to connect!
