Image Forgery Detection Using Deep Learning Framework
Yuchao Deng · 2022 IEEE 5th International Conference on Information Systems and Computer Aided Education (ICISCAE) · 2022
With the continuous development of digital media and the rapid change of digital image processing equipment, digital images are easy to be tampered with by criminals. In recent years, illegal elements often use tampering images to commit crimes, so it is urgent to study image forensics technology. Blind forensics technology is used to verify the original and authenticity of the image, and then the tampering mode of the image is discriminated and the tampered area is located according to the tampering trace of the image. This paper focuses on the detection of image manipulation and tampering and the localization of tampered areas. The main innovation of this paper is to use the core idea of end-to-end training of u-net network to further optimize the performance of image tampering detection tasks in combination with residual network.