A Novel Framework for Deepfake Image Detection Using Deep Learning Approach
Ambika Jaiswal, Aruna J. Chamatkar, Akash Prakash Kharat, Bhisham Sharma, Imed Ben Dhaou, Suhashini Awadhesh Chaurasia · 2025
Deepfake, fast growing field in the age of multimedia and AI, have been attracted attention in the recent years. Deep learning algorithms are used to create real digital content which are difficult to distinguish from authentic content. Deepfakes can serve multiples purposes, including educational content, academic research, social entertainment so it spread wrong information, information manipulation, damage to reputation, many frauds. Day to day deepfake crime is increasing. Detection of deepfake is big challenging issue in the digital forensics. A strong approach needs to be created from protecting against the media deepfake. The main work of this paper is to build a efficient framework for deepfake detection. The paper discusses deepfakes, detection techniques, datasets, proposed model. Novel proposed approach has been created that are based on the integration of CNN and VGG16. The deepfake dataset has been used to utilized the network architecture. Proposed model has been achieved the 95% accuracy and 94% precision score. The developed model is superior than the other state of art developed methods.