Deepfake Video Detection Using Lip Region Analysis with Advanced Artificial Intelligence Based Anomaly Detection Technique
The proliferation of internet usage has led to an increase in deepfake attacks, posing significant threats to privacy and data security. Existing detection systems are continually challenged by increasingly sophisticated deepfake techniques. In this paper, we propose a novel method for detecting deepfake anomalies by focusing on the lip region of human faces in videos. This area is often subtle and difficult for humans to scrutinize. Our approach integrates the Minimum Covariance Determinant (MCD) Estimator with the SHA-256 hashing algorithm and RAID technology to identify even the slightest deepfake activities. By employing the Lip Shaping Technique, we evaluate the effectiveness of our method. Experimental results demonstrate the proposed method’s promising performance and its significant impact on frame processing speed due to the incorporation of optimized storage techniques.