Authors: Prof. Keerti M, Mr.Narendra, Mr.Vishal, Mr.Kevin Dutt
Abstract: Deep fake detection technology has advanced rapidly with the progress of deep learning, enabling the generation of highly realistic manipulated images and videos. While such technology has beneficial applications in entertainment and media, its misuse poses serious threats including misinformation, identity fraud, political manipulation, and erosion of public trust. Traditional video authentication techniques are insufficient to detect subtle manipulations introduced by modern deepfake generation algorithms. This paper presents a deep learning–based deepfake detection system that analyzes both spatial and temporal inconsistencies in video frames. The proposed approach employs transfer learning–based convolutional neural networks for facial feature extraction and sequence-based models for capturing temporal variations across frames. Preprocessing techniques such as face detection, frame extraction, normalization, and data augmentation are applied to enhance detection robustness. Experimental evaluation using benchmark datasets demonstrates that the proposed system achieves reliable detection accuracy even for high-quality deepfake videos. The system provides an effective and scalable solution for digital forensics, cybersecurity, and social media content verification.
