Deep Detect

Deep Detect

Nirzara Manade1*, Shravani Shinde1, Sonal Patil1, Sunetra Bhambure1, S.V.Chavan 2

1Under Graduate Student, Department of Computer Science and Engineering, Sanjay Ghodawat Institute, Atigre, Kolhapur, Maharashtra, India

2HOD, Department of Computer Science and Engineering , Sanjay Ghodawat Institute, Atigre, Kolhapur, Maharashtra, India

*Corresponding Author: nirzaramanade@gmail.com

Abstract

The deepfake technology which has been evolved over the past few years, threatened the way our society perceives reality and might be a catalyst for disruptions on an ethical, social or security level. These skills at manipulating video and audio with precision open up a whole universe of potential risks ,misinformation, identity theft, privacy. In this paper it describes similar work in that it leverages writing style to detect deepfake content but it aims to develop a web-based application as an alternative way of detecting such malicious videos. The web-application, built on machine learning algorithms and deployed in an easily usable GUI  form factor, acts as a robust tool for detecting manipulations applied to media. It covers the key challenges like model accuracy, dataset quality & real-time processing various problematic areas are handled. From the results of implementation and tests, we will achieve some reasonable outcomes that demonstrate an easy way to cope up with them if it becomes more difficult in future times due to this rapidly increasing its technology and becoming more realistic day by day.

Keywords- Deep Fake, Machine Learning Algorithms, Identity theft, Real Time Processing, Web application