Discipline: Computer Sciences and Information Management
Subcategory: Computer Science & Information Systems
Session: 1
Room: Exhibit Hall A
Jahmelia Atkinson - Fort Valley State University
Co-Author(s): Peng Jiang
Smartphones are drastically evolving, they are equipped with the capabilities of performing intelligent activities like sensing, computing and networking. They allow users to perform a huge range of duties such as, socializing, communicating, storing and having access quick access to personal information. Today, majority of people have a tendency to store exclusive sorts of sensitive assets on their phone. These records may consist of personal Ids, bank account information and other valuable information; Because of this, smartphones are exposed to a slew of security threats and attacks from attackers. Smartphone authentication mechanisms are susceptible to numerous attacks such as smudge attacks, guessing, shoulder-smurfing. To tackle these challenges, we introduce continuous authentication of smartphone users. In this study, we examined how a smartphone user can be recognized based solely on their touchscreen interactions. We examined the touchscreen interactions of 8 different subjects performing distinctive touchscreen gestures. We have created an android application to capture the users unique touch events. For authentication purposes and recognizing a user we proposed a Convolutional Neural Network model using Keras in python. We anticipate this research will help significantly minimize breaches due to unauthorized access to smartphones.
Funder Acknowledgement(s): National Science Foundation and Department of Education
Faculty Advisor: Dr. Masoud Naghedolfeizi, feizim@fvsu.edu
Role: I carried out this research under the supervision of my advisors and with the help of my co-author.