Student, faculty receive award for research on deepfake media

Headshots of Mitra, Mohanty, and Kougianos

Doctoral student Alakananda Mitra, Computer Science and Engineering Professor Saraju P. Mohanty and Electrical Engineering Professor Elias Kougianos received the Best Paper Award at the OITS/IEEE International Conference on Information Technology, held in December.

The paper, titled, “Detection of Deep-Morphed Deepfake Images to Make Robust Automatic Facial Recognition Systems,” proposes a new method for detecting deepfake images. Deepfake media are digitally generated images and videos that use artificial intelligence and deep learning technologies.

"Deepfake media can have serious technical and social issues," said Mitra. "For example, disinformation using deepfake images or videos is possible. Deepfakes can also defeat AI systems."

Deepfake media poses a risk to Facial Recognition Systems used for security and access control in various military, technological, industrial and smart cities applications.

"The key idea behind the paper is the detection of these deepfake images in the context of Facial Recognition Systems of smart cities. By using a Convolutional Neural Network-based method, we can detect these deepfakes and process them faster, enhancing detection and security of Facial Recognition Systems," she said.

Mitra's research focuses on data quality in IoT-enabled systems. She has co-authored six articles and papers within the field and plans to study data quality aspects within smart cities and smart agriculture.

The Department of Computer Science and Engineering Smart Electronic Systems Laboratory (SESL), directed by Mohanty and Kougianos, has graduated 12 Ph.D. and 25 M.S. students. The SESL currently has 10 Ph.D. students engaged in cutting-edge research in IoT-enabled systems, smart healthcare and smart agriculture.