ANALYSIS ON SPLICING AND DEEPFAKES IN VISUAL FORGERY AND IMAGE FRAUD

Computer Science

Image forgery has become increasingly sophisticated, with deepfakes and GAN-generated images causing people to be more skeptical of visual evidence. This has a significant impact on politics, healthcare, terrorism, social media, and the legal system. As GANs and deepfake technologies evolve, the line between "real" and "fake" becomes blurrier, leading to new ethical, legal, and technical challenges. Society must balance the benefits of these technologies with the need for regulations and detection systems to combat their misuse. Splicing is the process of cutting and pasting parts of an image or video from one source to create an authentic composite, while deepfakes are synthetic media created using Generative Adversarial Networks (GANs) or Autoencoders. These forgeries generate new, realistic content, rather than copying and pasting it. Detecting these forgeries requires advanced deep learning models like CNNs, RNNs, LSTMs, and Transformers. Combating splicing and deepfakes is crucial in forensics, media verification, and financial fraud prevention. This paper reviews research work on visual media alteration to prevent frauds using deep learning methods, highlighting the importance of thorough knowledge of state-of-the-art approaches for detection

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.