Year: 2025
Type: article
Abstract: The rapid advancement of deep fake technology, particularly face-swap deep fakes, has introduced significant challenges to digital media integrity and cybersecurity. This paper proposes an AI/ML-based solution for detecting face-swap deep fake videos by utilizing Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) for spatial and temporal analysis. The system focuses on identifying inconsistencies across video frames, ensuring efficient and scalable detection even in resource-constrained environments. Initial results demonstrate the system's potential in addressing the growing threat of deep fake videos, with applications in media integrity and law enforcement.
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: bronze
OpenAlex ID: W4410178008