Deep learning algorithm for digital image forensics
Shagun Verma, Rahul Singh Chauhan, Ruchira Rawat, Pratibha · 2024
The creative population uses easy ways of editing and generating data, which has become easy to share and edit all over the Internet. Therefore, it has become crucial to verify the authenticity and to use more innovative methods to avoid problems like copy paste and deepfake. There are a number of methods that can help people come out of these problems in a minute, like using deep learning techniques to analyze images practically. These techniques would help us to separate out the real and artificially generated contents, or abstracts. Nowadays, deep learning is the most technical and intellectual concept to overcome from deepfake. This survey demonstrates how deep learning can help distinguish between real and fraudulent images: Through this technique, it provides reliability for creating and evolving it. This method allows us to examine the advantages and disadvantages of the above model with respect to its limitations as well. Assuming this article represents all the awaited potential to give a correct direction to the development and innovation that helps others. It helps people and society in a way to input their problem, process it, and give a useful output. This comprehensive guide is made to support image forgery and its detection and to provide rigid boundary against proliferation of contents.