A Comparative Study of Deep Learning Techniques for Deepfake Detection

Abstract

Deepfake media generated or altered using artificial intelligence can imitate real people and events with increasing visual realism, creating challenges for digital trust, privacy, cybersecurity, journalism, and forensic investigation. This paper presents a comparative literature-based study of deep learning techniques used for deepfake detection. Representative approaches are examined across spatial, temporal, frequency-aware, transformer-based, and hybrid categories. Major public datasets, including FaceForensics++, Celeb-DF, and the DeepFake Detection Challenge dataset, are also compared. The analysis shows that CNN-based methods provide efficient spatial feature extraction, temporal models capture frame-to-frame inconsistencies, frequency-aware methods expose complementary forensic traces, and transformer-based methods offer broader contextual modelling at higher computational cost. Recent benchmark and generalization studies indicate that performance measured only within the training distribution can overstate practical reliability. The paper therefore emphasizes cross-dataset evaluation, robustness to compression, computational efficiency, explainability, and reproducibility as important criteria for practical deepfake detection