Multimedia Forensics Before the Deep Learning Era

Abstract Image manipulation is as old photography itself, and powerful media editing tools have been around for a long time. Using such conventional signal processing methods, it possible to modify images videos obtaining very realistic results. This chapter devoted describe the most effective strategies detect widespread manipulations that rely on traditional approaches do not require deep learning strategy. In particular, we will focus like adding, replicating, or removing objects present major lines of research in multimedia forensics before era rise deepfakes. The popular look artifacts related in-camera chain (camera-based clues) out-camera history (editing-based clues). We methods extraction camera fingerprint need some prior information pristine data, example, through collection taken from interest. Then shift blind any knowledge reveal inconsistencies with respect well-defined hypotheses. also briefly review interesting features machine learning- based finally challenges this area.