Detecting Deep fake Videos Using Advanced Deep Learning Techniques

Abstract

Deep fakes have recently exploded in popularity; they are films and photographs that have been digitally changed to seem quite genuine. Researchers are looking at a lot of amazing applications for this technology. There has been a recent uptick in the fraudulent usage of films online, including those promoting false news, pornographic movies featuring celebrities, and financial schemes. Therefore, the Deep fake detection problem is especially harmful to politicians, celebrities, and other prominent people. Many algorithms based on deep learning have been proposed to identify deep fake films or photographs, and a great deal of study has been conducted in the last few years to comprehend the inner workings of deep fakes. Research in this area has focused on deep fake creation and detection systems that use various deep learning techniques. Also covered will be the limitations of existing methods as well as the societal availability of databases. The development of an automated, deep fake detection system. The world faces a major challenge due to the absence of an efficient deep fake detection system, since deep fake films and pictures may be easily made and circulated. Still, many have tried to solve this problem, and the ones that include deep learning have shown to be more effective than the more conventional methods. With these features, we can train ResNext to detect when a video has been manipulated and when it hasn't, as well as to spot the time discrepancies between frames shown by DF introduction tools.