Robust Texture-Aware Computer-Generated Image Forensic: Benchmark and Algorithm

With advances in rendering techniques and generative adversarial networks, computer-generated (CG) images tend to be indistinguishable from photographic (PG) images. Revisiting previous works towards CG image forensic, we observed that existing datasets are constructed years ago and limited in both quantity and diversity. Besides, current algorithms only consider the global visual features for forensic, ignoring finer differences between CG and PG images. To mitigate these problems, we first contribute a Large-Scale CG images Benchmark (LSCGB), and then propose a simple yet strong baseline model to address the forensic task. On the one hand, the introduced benchmark has three superior properties, 1) large-scale: the benchmark contains 71168 CG and 71168 PG images with the corresponding expert-annotated labels. It is orders of magnitude bigger than previous datasets. 2) high diversity: we collect CG images from 4 different scenes generated by various rendering techniques. The PG images are varied in terms of image content, camera models, and photographer styles. 3) small bias: we carefully filter the collected images to ensure that the distributions of color, brightness, tone and saturation between CG and PG images are close. Furthermore, inspired by an empirical study on texture difference between CG and PG images, an effective texture-aware network is proposed to improve forensic accuracy. Concretely, we first strengthen texture information of multilevel features extracted from a backbone. Then, the relations among feature channels are explored by learning its gram matrix. Each feature channel represents a specific texture pattern. The gram matrix is thus able to embed the finer texture differences. Experimental results demonstrate that this baseline surpasses the existing methods. The benchmark is publically available at https://github.com/wmbai/LSCGB.

Image Forensics

From mainstream media outlets to social media and everything in between, doctored photographs are appearing with growing frequency and sophistication. The resulting lack of trust is impacting law enforcement, national security, the media, e-commerce, and more. While some types of manipulations can be detected with a careful visual examination, our visual system seems unable to reliably detect other types of manipulations. The field of image forensics has emerged to help return some trust in photography. I describe the perceptual limits of detecting manipulated images, as well as representative examples of computational techniques for authenticating images.

Identifying Computer-Generated Portraits: The Importance of Training and Incentives

The past two decades have seen remarkable advances in photo-realistic rendering of everything from inanimate objects to landscapes, animals, and humans. We previously showed that despite these tremendous advances, human observers remain fairly good at distinguishing computer-generated from photographic images. Building on these results, we describe a series of follow-up experiments that reveal how to improve observer performance. Of general interest to anyone performing psychophysical studies on Mechanical Turk or similar platforms, we find that observer performance can be significantly improved with the proper incentives.

Estimate of PRNU Noise Based on Different Noise Models for Source Camera Identification

Identification of the source that has generated a digital content is considered one of the main open issues in multimedia forensics community. The extraction of photo-response non-uniformity (PRNU) noise has been so far indicated as a mean to identify sensor fingerprint. Such a fingerprint can be estimated from multiple images taken by the same camera by means of a de-noising filtering operation. In this paper, the authors propose a novel method for estimating the PRNU noise in source camera identification. In particular, a MMSE digital filter in the un-decimated wavelet domain, based on a signal-dependent noise model, is introduced and compared with others commonly adopted for this purpose. A theoretical framework and experimental results are provided and discussed.

A DFT-Based Analysis to Discern Between Camera and Scanned Images

Digital images are generated by different sensors, understanding which kind of sensor has acquired a certain image could be crucial in many application scenarios where digital forensic techniques operate. In this article a new methodology which permits to establish if a digital photo has been taken by a photo-camera or has been scanned by a scanner is presented. The specific geometrical features of the sensor pattern noise introduced by the sensor are investigated by resorting to a DFT (Discrete Fourier Transform) analysis and consequently the origin of the digital content is assessed. Experimental results are provided to witness the reliability of the proposed technique.