Students who are considering working on the following project topics for their graduation design project can contact Prof. Dr. Mehmet Keskinöz for detailed information.

E-mail:
keskinoz@itu.edu.tr
Office No: BBF-3307


Detecting Social Bots Using Twitter Data
Social bots have been very active in online spaces for marketing, political campaigns, or emotional manipulation. In this project, we are aiming at developing techniques to detect twitter bots with high accuracy.
For this project, one of the aims is to develop a browser extension, which gathers twitter data and sends it to a machine learning platform and the prediction result of the machine learning platform will be displayed on a website to improve the user experience.


Development of A Secure Biometric Authentication System
In this project, the students are required to develop a simple and secure biometric (e.g, person’s face image, voice , palmprint etc.) authentication system.  In a generic scenario, biometric information is stored in a smart card  or biometric RF-ID tags for authentication purposes. If the card is lost or stolen, there is a big security issue.  To alleviate this problem, students will implement encrypted biometric templates.


Multimedia Security Through Digital Watermarking
Rapid development of Internet has greatly increased the need for creation, storage and distribution of digital multimedia products. This raises, however, security concerns due to digital multimedia products high vulnerability to the illegal copying, distribution, manipulation, and other attacks. To remedy these security issues, in literature, the idea of “the digital watermarking” has been developed where the information to be hidden is carried by the watermark signal that is transmitted over the host signal.
In this project, students will investigate the popular digital  image watermarking techniques and develop a graphical user interface that visually illustrates their work.


Privacy Preserving Face Tracking Through Data Anonymization
In this project, students will first employ deep learning techniques for face anonymization  and track the face of a person through his/her anonymized face recordings.