DEEP FAKE DETECTION

DEEP FAKE DETECTION

Ms. Santhoshi P

Anupama R,   Harish J, Krish Joyeaal A B

BACHELOR OF TECHNOLOGY – 3rd YEAR

DEPARTMENT OF ARITIFICIAL INTELLIGENCE AND DATA SCIENCE

SRI SHAKTHI OF ENGINEERING AND TECHNOLOGY(AUTONOMOUS) COIMBATORE-641062

ABSTRACT
Artificial intelligence has led to many improvements, but it is coupled with its own issues, such as deepfakes. Videos, images, and audio can be synthesized using advanced deep learning methods, such as GANs, leading to the rise of deepfakes that creates fake but realistic looking images, videos, dressed voices, etc. even to the extent of realness where one cannot tell whether a media is real or fake. It is a helpful technology made to inspire creativity and to entertain people, but it has been abused for the last decade, especially for malicious purposes like propaganda, impersonation, bullying, to the extent of threatening people's privacy, safety, and belief in the sanity of digital content. Here within, review of deepfake detection approaches is presented, focusing on the very latest techniques and how they might address the challenges presented by these technologies.

Keywords: Deepfakes, GANs, privacy, detection, CNNs, RNNs, artifacts, facial movements, synthetic media, ethical risks.