Deep Fake Detection System

Deep Fake Detection System

Author Name : Dhanashree Shelke, Shruti Patil, Juhi Raut, Prof. Shudhodhan Bokefode, Akshata Dange

ABSTRACT :  With the proliferation of deepfake videos posing significant challenges in misinformation and manipulation, the necessity for reliable detection methods has become paramount. This abstract investigates the utilization of Long Short Term Memory (LSTM) networks within the domain of deepfake video detection. LSTM, a form of recurrent neural network (RNN), exhibits proficiency in capturing temporaldependencies within sequential data, thereby presenting a promising avenue for analyzing the dynamic characteristics of videos. The research explores the nuances of employing LSTM architectures for discerning deepfake videos, underscoring the importance of comprehending temporal patterns intrinsic to manipulated content.The findings of this research have important practical ramifications, particularly for social networking and video hosting services. The incorporation of deepfake detection based on LSTM technology has the potential to promote an online environment that is more secure and safe.