Abstract

The rise of deepfakes, realistic, AI-generated videos and images, has brought challenges to the integrity of digital media and information in today's society. From political manipulation to social media disinformation, deepfakes have weakened public trust and sparked concerns about privacy, security, and the ethical use of AI technologies. This paper explores the role of deepfake detection software. We begin by providing a comprehensive overview of what deepfakes are, their machine learning principles and the architecture of CNNs.The study then investigates various activation function approaches to deepfake detection, evaluating the performance and limitations of current models. This thesis hopes to contribute to the ongoing discussion on how these types of software can be harnessed to protect the integrity of digital media along with the harmful effects of digital misinformation.

Advisor

As'ad, Asa'd

Department

Computer Science

Disciplines

Cybersecurity | Numerical Analysis and Scientific Computing

Publication Date

2025

Degree Granted

Bachelor of Arts

Document Type

Senior Independent Study Thesis

Share

COinS
 

© Copyright 2025 Charles B. Carroll III