A Comprehensive Review of Deep Learning Techniques for Image Forgery Detection

The rapid growth of multimedia content and advanced editing tools has led to widespread manipulation of digital images. While image editing serves artistic and legitimate purposes, it also presents significant risks, including misinformation, fraud and cybercrimes. As a result, image forgery detection has become a critical research area aimed at preserving the authenticity and reliability of visual content, especially in fields like journalism, law enforcement and digital forensics. Image forgery techniques have evolved from basic manipulations, such as splicing and copy-move, to more advanced methods like deepfakes, which use deep learning to generate hyper-realistic altered visuals. These challenges demand the development of sophisticated detection methods capable of adapting to these advanced forgeries.