ABSTRACT
Software packages such as Autodesk Maya and Blender have given a significant lift to the photorealism industry which aims at camouflaging computer-generated photos as photographic images (PIM). Although an asset for the movies and video game industry and elsewhere also, with advancements in photorealism, images and videos have lost the stature of a genuine record of memories and facts. Moreover, from the security perspectives, this raises serious concerns about forging evidence. Consequently, in this chapter, we propose a system for the automatic detection of photorealistic computer-generated images (PRCG). The proposed system utilizes rich models for differentiating PRCG from photographic images (PIM). The proposed system achieves close to 99% classification accuracy in many scenarios.
