Detecting "DeepFakes" in H.264 Video Data Using Compression Ghost Artifacts

In recent years, the number of forged videos circulating onthe Internet has immensely increased. Software and services tocreate such forgeries have become more and more accessible tothe public. In this regard, the risk of malicious use of forgedvideos has risen. This work proposes an approach based on theGhost effect knwon from image forensics for detecting forgeriesin videos that can replace faces in video sequences or change themimic of a face. The experimental results show that the proposedapproach is able to identify forgery in high-quality encoded videocontent.