Advancements in Deepfake Detection: A Comprehensive Review of AI-Driven Approaches
L. Khumanleima Devi, B Ben Sujitha · 2025
The significant developments in artificial neural network (ANN)-based technologies are crucial for manipulating multimedia content. Although technology has predominantly been used for purposes, such as education, some individuals have exploited it for unlawful or harmful activities. This literature review offers a comprehensive analysis of the advancements in deepfake detection. This study explores state-of-the-art approaches, including convolutional neural network (CNNs), transformers, ensemble model, and binary neural network, that leverage both spatial and temporal feature for effective detection. In addition, the review examines the role of data augmentation, artifact attention, and multi-modal analysis in enhancing detection accuracy across diverse datasets. By analyzing current methodologies and evaluating their strengths and limitations, this paper provides insights into the evolving landscape of deepfake detection by analyzing current methodologies and highlights key challenges such as the need for robust datasets and cross-domain generalization. Future directions are also outlined, emphasizing the potential of self-supervised learning, ensemble approaches, and advancements in artifact detection for more reliable and scalable solutions.