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
Recommended Citation
Carroll, Charles B. III, "Activation Functions in Deepfake Detection" (2025). Senior Independent Study Theses. Paper 11424.
https://openworks.wooster.edu/independentstudy/11424
Disciplines
Cybersecurity | Numerical Analysis and Scientific Computing
Publication Date
2025
Degree Granted
Bachelor of Arts
Document Type
Senior Independent Study Thesis
© Copyright 2025 Charles B. Carroll III