DeepFake Creation and Detection:A Survey

DeepFake has become very popular of late. The term DeepFake refers to any multimedia content generated using deep learning technology appearing realistic to people. Despite the beneficial advances of DeepFake, it has been a major cause of threats to a person's privacy, where one person's face can be swapped with another in an indistinguishable way without consent. Also, it is easy for malicious parties to take over public events like elections by spreading misinformation and leaving a negative impact on national security. Thus, detection of such DeepFake is a crucial yet challenging problem. Human-eye-based segregation of DeepFaked contents from real ones has always been a difficult task; but recent works have shown the use of different technologies recording good results for the same, although with some limitations. The paper thus explores different algorithms used for DeepFake creation and detection; presenting a comprehensive overview of the techniques used and aimed at identifying their pros and cons.