A Review on Deepfake Image Detection Approaches, Techniques and Methods

Deepfake technology has been increasingly advanced and highly capable of generating and modifying media contents. Deepfake technologies like automated generation of scripts and avatars creation are applicable on many distinct fields and industries, helping in improving and enhancing the visuals and media contents to be published and shared with the public. However, Deepfake technologies also, being widely used in various malicious activities and crime cases, like financial frauds and scams, as well as sexual abuse and harassment. Hence, Deepfake detection is essential, to reduce the problems and avoid the consequences brought by Deepfake technologies. In this work, several existing Deepfake detection approaches, techniques and methods are explored and discussed. To understand better the effects and consequences caused by Deepfake, this work includes cases studies on real-world Deepfake incidents, conducted by searching through the news information accessible through internet. In addition, a thorough survey is conducted by searching for related articles and papers via Google Scholar using a suitable set of keywords. Results are presented via taxonomy. Some comparisons are presented to give some insights on the deepfake image detection approaches.