Andrea Costanzo

Exploring Image Dependencies: a New Challenge in Image Forensics

Though the current state of the art of image forensics permits to acquire very interesting information about image history, all the instruments developed so far focus on the analysis of single images. It is the aim of this paper to propose a new approach that moves the forensics analysis further, by considering groups of images instead of single images. The idea is to discover dependencies among a group of images representing similar or equal contents in order to construct a graph describing image relationships. Given the pronounced effect that images posted on the Web have on opinions and bias in the networked age we live in, such an analysis could be extremely useful for understanding the role of pictures in the opinion forming process. We propose a theoretical framework for the analysis of image dependencies and describe a simple system putting the theoretical principles in practice. The performance of the proposed system are evaluated on a few practical examples involving both images created and processed in a controlled way, and images downloaded from the web.

Identification of cut & paste tampering by means of double-JPEG detection and image segmentation

This paper focuses on images whose content has been modified by means of a cut & paste operation. By relying on an existing scheme for the detection of double JPEG compressed images with desynchronized grids, we propose two algorithms for the detection of image regions that have been transplanted from another image. The proposed methods work whenever the pasted region is extracted from a JPEG compressed image and inserted in a target image that is subsequently compressed with a quality factor larger than that used to compress the source image. The new methods are intended as a complement to previous works relying on the detection of artifacts introduced by double JPEG compression with aligned compression grids. The experiments we carried out show the good performance of the novel schemes, the second one providing better results at a lower complexity thanks to the incorporation within the detection process of some information regarding the actual image content.

Forensic Analysis of SIFT Keypoint Removal and Injection

Attacks capable of removing SIFT keypoints from images have been recently devised with the intention of compromising the correct functioning of SIFT-based copy-move forgery detection. To tackle with these attacks, we propose three novel forensic detectors for the identification of images whose SIFT keypoints have been globally or locally removed. The detectors look for inconsistencies like the absence or anomalous distribution of keypoints within textured image regions. We first validate the methods on state-of-the-art keypoint removal techniques, then we further assess their robustness by devising a counter-forensic attack injecting fake SIFT keypoints in the attempt to cover the traces of removal. We apply the detectors to a practical image forensic scenario of SIFT-based copy-move forgery detection, assuming the presence of a counterfeiter who resorts to keypoint removal and injection to create copy-move forgeries that successfully elude SIFT-based detectors but are in turn exposed by the newly proposed tools.

Counter-forensics of SIFT-based copy-move detection by means of keypoint classification

Copy-move forgeries are very common image manipulations that are often carried out with malicious intents. Among the techniques devised by the ‘Image Forensic’ community, those relying on scale invariant feature transform (SIFT) features are the most effective ones. In this paper, we approach the copy-move scenario from the perspective of an attacker whose goal is to remove such features. The attacks conceived so far against SIFT-based forensic techniques implicitly assume that all SIFT keypoints have similar properties. On the contrary, we base our attacking strategy on the observation that it is possible to classify them in different typologies. Also, one may devise attacks tailored to each specific SIFT class, thus improving the performance in terms of removal rate and visual quality. To validate our ideas, we propose to use a SIFT classification scheme based on the gray scale histogram of the neighborhood of SIFT keypoints. Once the classification is performed, we then attack the different classes by means of class-specific methods. Our experiments lead to three interesting results: (1) there is a significant advantage in using SIFT classification, (2) the classification-based attack is robust against different SIFT implementations, and (3) we are able to impair a state-of-the-art SIFT-based copy-move detector in realistic cases.

A fuzzy approach to deal with uncertainty in image forensics

Image forensics research has mainly focused on the detection of artifacts introduced by a single processing tool, thus resulting in the development of a large number of specialized algorithms looking for one or more specific footprints under precise settings. As one may guess, the performance of such algorithms are not ideal, so the output they provide may be noisy, inaccurate and only partially true. Moreover, in real scenarios a manipulated image is often the result of the application of several tools made available by the image processing software. As a consequence, reliable tamper detection requires that several tools developed to deal with different scenarios are applied. The above observations raise two new problems: (i) deal with the uncertainty introduced by error-prone tools and (ii) devise a sound strategy to merge the information provided by the different tools into a single output. To overcome these problems we propose a decision fusion framework based on the Fuzzy Theory, which permits to cope with the uncertainty and lack of precise information typical of image forensics, by leveraging on the widely known ability of the Fuzzy Theory to deal with inaccurate and incomplete information. We describe a practical implementation of the proposed framework and validate it in a realistic scenario in which five forensic tools exploit JPEG compression artifacts to detect cut&paste tampering within a specified region of an image. The results are encouraging, and provide a significant advantage with respect to those obtained by simply OR-ing the outputs of the single tools.

On the effectiveness of local warping against SIFT-based copy-move detection

One of the simpler and most used method to alter the content of a digital image is to copy-move a portion of it onto another area with the intent, usually, to hide something awkward. In image forensics scientific community, this kind of modification is generally detected by resorting at techniques based on SIFT features that provide a local description which is robust to global geometric transformations the image may undergo. On such a basis, this paper investigates the effectiveness of some methodologies which introduce a local warping onto the copy-pasted patches in order to reduce the detection capability of SIFT-based approaches. This analysis is particularly interesting in a real scenario of forensic security. Four diverse local warping techniques have been taken into account and experimental results with respect to final perceptual quality of the forged image are presented.

Dealing with uncertainty in image forensics: A fuzzy approach

Image forensics research has mainly focused on the detection of artifacts introduced by a single processing tool. In tamper detection applications, however, the kind of artifacts the forensic analyst should look for is not known beforehand, hence making it necessary that several tools developed for different scenarios are applied. The problem, then, is twofold: i) devise a sound strategy to elaborate the information provided by the different tools into a single output, and ii) deal with the uncertainty introduced by error-prone tools. In this paper, we introduce a framework based on Fuzzy Theory to overcome these problems. We describe a practical implementation of the proposed framework putting the theoretical principles in practice. To validate the proposed approach, we carried out some experiments addressing a simple realistic scenario in which three forensic tools exploit artifacts introduced by JPEG compression to detect cut&paste tampering within a specified region of an image. The results are encouraging, especially when compared with those obtained by simply XOR-ing the output of the the single detection tools.

Removal and injection of keypoints for SIFT-based copy-move counter-forensics

Recent studies exposed the weaknesses of scale-invariant feature transform (SIFT)-based analysis by removing keypoints without significantly deteriorating the visual quality of the counterfeited image. As a consequence, an attacker can leverage on such weaknesses to impair or directly bypass with alarming efficacy some applications that rely on SIFT. In this paper, we further investigate this topic by addressing the dual problem of keypoint removal, i.e., the injection of fake SIFT keypoints in an image whose authentic keypoints have been previously deleted. Our interest stemmed from the consideration that an image with too few keypoints is per se a clue of counterfeit, which can be used by the forensic analyst to reveal the removal attack. Therefore, we analyse five injection tools reducing the perceptibility of keypoint removal and compare them experimentally. The results are encouraging and show that injection is feasible without causing a successive detection at SIFT matching level. To demonstrate the practical effectiveness of our procedure, we apply the best performing tool to create a forensically undetectable copy-move forgery, whereby traces of keypoint removal are hidden by means of keypoint injection.

Attacking image classification based on bag-of-visual-words

Nowadays, with the widespread diffusion of online image databases, the possibility of easily searching, browsing and filtering image content is more than an urge. Typically, this operation is made possible thanks to the use of tags, i.e., textual representations of semantic concepts associated to the images. The tagging process is either performed by users, who manually label the images, or by automatic image classifiers, so as to reach a broader coverage. Typically, these methods rely on the extraction of local descriptors (e.g., SIFT, SURF, HOG, etc.), the construction of a suitable feature-based representation (e.g., bag-of-visual words), and the use of supervised classifiers (e.g., SVM). In this paper, we show that such a classification procedure can be attacked by a malicious user, who might be interested in altering the tags automatically suggested by the classifier. This might be used, for example, by an attacker who is willing to avoid the automatic detection of improper material in a parental control system. More specifically, we show that it is possible to modify an image in order to have it associated to the wrong class, without perceptually affecting the image visual quality. The proposed method is validated against a well known image dataset, and results prove to be promising, highlighting the need to jointly study the problem from the standpoint of both the analyst and the attacker.

SIFT keypoint removal and injection for countering matching-based image forensics

Scale Invariant Feature Transform (SIFT) has been widely employed in several image application domains, including Image Forensics (e.g. detection of copy-move forgery or near duplicates). Until now, the research community has focused on studying the robustness of SIFT against legitimate image processing, but rarely concerned itself with the problem of SIFT security against malicious procedures. Recently, a number of methods allowing to remove SIFT keypoints from an original image have been devised. Although quite effective, such methods produce an attacked image with very few (or no) keypoints, thus leaving cues that can be easily exploited by a forensic analyst to reveal the occurred manipulation. In this paper, we explore the topic of reintroducing fake SIFT keypoints into a previously cleaned image in order to address the main weakness of the existing removal attacks. In particular, we evaluate the fitness of locally adaptive contrast enhancement methods to the task of injecting new keypoints. The results we obtained are encouraging: (i) it is possible to effectively introduce new keypoints whose descriptors do not match with those of the original image, thus concealing the removal forgery; (ii) the perceptual quality of the image following the removal and injection attacks is comparable to the one of the original image.