A Review on Defake Detection Technique for Human Faces

This paper presents DeepDetect, an innovative and clearly explainable hybridized system for detecting deepfakes that can accurately differentiate between authentic image files (e.g. photographs) and computer-generated picture files, also known as fake or artificial pictures. Unlike previous approaches that have relied heavily on Convolutional Neural Networks (CNNs), DeepDetect does not require this methodology to achieve its results. Rather, it combines digital forensics methods for extracting features from digital media files with statistical approaches to machine learning using XGBoost and ensemble methods to make rigorous decisions about images based on a variety of criteria while also being interpretable. Forensic indicators are utilized to detect the characteristics of fake images that are not easily concealable by generative models. Statistical and physical model evidence are both leveraged together to produce more robust decisions and better serve the user. The framework, which uses ReactJS as the front end and FastAPI, OpenCV, and MongoDB as the backend technology stack, will enable users to analyzing images quickly and efficiently. Overall, DeepDetect will provide a viable and dependable method of detecting deepfakes across a variety of digital forensic, media authentication, and cybersecurity applications.

[1]   A. Rossler, D. Cozzolino, L. Verdoliva, C. Riess, J. Thies and M. Nießner, “FaceForensics++: Learning to Detect Manipulated Facial Images,” in Proc. IEEE/CVF Int. [2]   F. Matern, C. Riess and M. Stamminger, “Exploiting Visual Artifacts to Expose Deepfakes and Face Manipulations,” in 2019 IEEE Winter Applications of Computer Vision Workshops (WACVW), 2019, pp. 83–92. [3]   L. Guarnera, O. Giudice and S. Battiato, “Fighting Deepfake by Exposing the Convolutional Traces on Images,” arXiv preprint arXiv:2008.04095, 2020.  [4]   G. Gupta, Y. Chen and S. Bhatia, “A Comprehensive Review of Deepfake Detection Using Machine Learning and Hybrid Techniques,” Electronics, vol. 13, no. 1, pp. 95, 2024.  [5]   M. Alrashoud, A. S. Malik and M. Khan, “Deepfake Video Detection Methods, Approaches, and Challenges: A Survey,” Alexandria Engineering Journal, vol. 64, no. 4, pp. 3121–3138, 2025.  [6]   D. L. T. Bale, L. C. Ochei and C. Ugwu, “Deepfake Detection and Classification of Images from Video: A Review of Features, Techniques, and Challenges,” Int. J. Intelligent Information Systems, vol. 13, no. 2, pp. 20–28, 2024. [7]   A. T. Nguyen, J. Y. J. Su and B. Kim, “Deepfakes Generation and Detection: State-of-the-Art, Open Challenges, Countermeasures, and Way Forward,” arXiv preprint arXiv:2103.00484, 2021.  [8]   D. J. Amerini, L. Galteri, R. Caldelli and A. Del Bimbo, “Deepfake Video Detection Through Optical Flow Based CNN,” in Proc. IEEE/CVF Int. Conf. on Computer Vision (ICCV), 2019.  [9]   T. Chen and C. Guestrin, “XGBoost: A Scalable Tree Boosting System,” in Proc. 22nd ACM SIGKDD Int. Conf. on Knowledge Discovery and Data Mining, 2016, pp. 785–794.  [10]   B. Dolhansky, J. Bitton, B. Pflaum, J. Lu and C. C. Ferrer, “The Deepfake Detection Challenge Dataset,” arXiv preprint arXiv:2006.07397, 2020.