DeepFake Detection Model using LSTM-CNN with ...

DeepFake Detection Model using LSTM-CNN with Image and Temporal Video Analysis

Author Name : Prof. Minal Jungare, Prisha Ganganmale , Rutuja Khandagale, Sakshi Dhamane, Anjali Susar

ABSTRACT: In today's digital landscape, the widespread use of deepfake technology presents a significant challenge to theauthenticity and trustworthiness of visual content. This project seeks to address this pressing issue by focusing on thedetection of deepfakes in both images and videos. Leveraging advanced machine learning techniques, including Convolutional Neural Networks (CNNs), the project aims to train models to discern characteristic patterns and anomalies indicative of manipulation. By exposing these models to diverse datasets containing both authentic and deepfake content, they learn to differentiate between genuine and fraudulent media. The approach involves robust analysis of individual frames in images and sequential frames in videos, allowing for comprehensive scrutiny of potential inconsistencies or artifacts introduced during the manipulation process. This versatile solution offers a means to combat fraudulent content effectively, safeguarding digital trustworthiness in an era where the authenticity of visual media is increasingly at risk.