Photo-response Non-uniformity Pattern Research Articles

Detection of AI-Created Images Using Pixel-Wise Feature Extraction and Convolutional Neural Networks.

Generative AI has gained enormous interest nowadays due to new applications like ChatGPT, DALL E, Stable Diffusion, and Deepfake. In particular, DALL E, Stable Diffusion, and others (Adobe Firefly, ImagineArt, etc.) can create images from a text prompt and are even able to create photorealistic images. Due to this fact, intense research has been performed to create new image forensics applications able to distinguish between real captured images and videos and artificial ones. Detecting forgeries made with Deepfake is one of the most researched issues. This paper is about another kind of forgery detection. The purpose of this research is to detect photorealistic AI-created images versus real photos coming from a physical camera. Id est, making a binary decision over an image, asking whether it is artificially or naturally created. Artificial images do not need to try to represent any real object, person, or place. For this purpose, techniques that perform a pixel-level feature extraction are used. The first one is Photo Response Non-Uniformity (PRNU). PRNU is a special noise due to imperfections on the camera sensor that is used for source camera identification. The underlying idea is that AI images will have a different PRNU pattern. The second one is error level analysis (ELA). This is another type of feature extraction traditionally used for detecting image editing. ELA is being used nowadays by photographers for the manual detection of AI-created images. Both kinds of features are used to train convolutional neural networks to differentiate between AI images and real photographs. Good results are obtained, achieving accuracy rates of over 95%. Both extraction methods are carefully assessed by computing precision/recall and F1-score measurements.

Experiencing with electronic image stabilization and PRNU through scene content image registration

• Two scene-based methods to estimate transformations introduced by EIS are presented. • The first exploits scene alignment to register different acquisition modalities. • The second uses a laboratory setup to assess the transformation used during EIS. • Outcomes can be exploited to improve and speed-up PRNU matching on stabilized videos. • Additionally, best practices for data acquisition are also discussed. This paper explores content-based image registration as a means of dealing with and understanding better Electronic Image Stabilization (EIS) in the context of Photo Response Non-Uniformity (PRNU) alignment. A novel and robust solution to extrapolate the transformation relating the different image output formats for a given device model is proposed. This general approach can be adapted to specifically extract the scale factor (and, when appropriate, the translation) so as to align native resolution images to video frames, with or without EIS on, and proceed to compare PRNU patterns. Comparative evaluations show that the proposed approach outperforms those based on brute-force and particle swarm optimization in terms of reliability, accuracy and speed. Furthermore, a tracking system able to revert back EIS in controlled environments is designed. This allows one to investigate the differences between the existing EIS implementations. The additional knowledge thus acquired can be exploited and integrated in order to design and implement better future PRNU pattern alignment methods, aware of EIS and suitable for video source identification in multimedia forensics applications.

Factors that Influence PRNU-Based Camera-Identification via Videos.

The Photo Response Non-Uniformity pattern (PRNU-pattern) can be used to identify the source of images or to indicate whether images have been made with the same camera. This pattern is also recognized as the “fingerprint” of a camera since it is a highly characteristic feature. However, this pattern, identically to a real fingerprint, is sensitive to many different influences, e.g., the influence of camera settings. In this study, several previously investigated factors were noted, after which three were selected for further investigation. The computation and comparison methods are evaluated under variation of the following factors: resolution, length of the video and compression. For all three studies, images were taken with a single iPhone 6. It was found that a higher resolution ensures a more reliable comparison, and that the length of a (reference) video should always be as high as possible to gain a better PRNU-pattern. It also became clear that compression (i.e., in this study the compression that Snapchat uses) has a negative effect on the correlation value. Therefore, it was found that many different factors play a part when comparing videos. Due to the large amount of controllable and non-controllable factors that influence the PRNU-pattern, it is of great importance that further research is carried out to gain clarity on the individual influences that factors exert.

PRNU‐based detection of facial retouching

Nowadays, many facial images are acquired using smart phones. To ensure the best outcome, users frequently retouch these images before sharing them, e.g. via social media. Modifications resulting from used retouching algorithms might be a challenge for face recognition technologies. Towards deploying robust face recognition as well as enforcing anti-photoshop legislations, a reliable detection of retouched face images is needed. In this work, the effects of facial retouching on face recognition are investigated. A qualitative assessment of 32 beautification apps is conducted. Based on this assessment five apps are chosen which are used to create a database of 800 beautified face images. Biometric performance is measured before and after retouching using a commercial face recognition system. Subsequently, a retouching detection system based on the analysis of photo response non-uniformity (PRNU) is presented. Specifically, scores obtained from analysing spatial and spectral features extracted from PRNU patterns across image cells are fused. In a scenario, in which unaltered bona fide images are compressed to the average sizes of the retouched images using JPEG, the proposed PRNU-based detection scheme is shown to robustly distinguish between bona fide and retouched images achieving an average detection equal error rate of 13.7% across all retouching algorithms.

Mitigation of H.264 and H.265 Video Compression for Reliable PRNU Estimation

The photo-response non-uniformity (PRNU) is a distinctive image sensor characteristic, and an imaging device inadvertently introduces its sensor's PRNU into all media it captures. Therefore, the PRNU can be regarded as a camera fingerprint and used for source attribution. The imaging pipeline in a camera, however, involves various processing steps that are detrimental to PRNU estimation. In the context of photographic images, these challenges are successfully addressed and the method for estimating a sensor's PRNU pattern is well established. However, various additional challenges related to generation of videos remain largely untackled. With this perspective, this work introduces methods to mitigate disruptive effects of widely deployed H.264 and H.265 video compression standards on PRNU estimation. Our approach involves an intervention in the decoding process to eliminate a filtering procedure applied at the decoder to reduce blockiness. It also utilizes decoding parameters to develop a weighting scheme and adjust the contribution of video frames at the macroblock level to PRNU estimation process. Results obtained on videos captured by 28 cameras show that our approach increases the PRNU matching metric up to more than five times over the conventional estimation method tailored for photos.

Detection of Face Morphing Attacks Based on PRNU Analysis

Recent research found that attacks based on morphed face images, i.e., morphing attacks, pose a severe security risk to face recognition systems. A reliable morphing attack detection from a single face image remains a research challenge since cameras and morphing techniques used by an attacker are unknown at the time of classification. These issues are commonly overseen while many researchers report encouraging detection performance for training and testing morphing attack detection schemes on images obtained from a single face database employing a single morphing algorithm. In this work, a morphing attack detection system based on the analysis of Photo Response Non-Uniformity (PRNU) is presented. More specifically, spatial and spectral features extracted from PRNU patterns across image cells are analyzed. Differences of these features for bona fide and morphed images are estimated during a threshold-selection stage using the Dresden image database which is specifically built for PRNU analysis in digital image forensics. Cross-database evaluations are then conducted employing an ICAO compliant subset of the FRGCv2 database and a Print-Scan database which is a printed and scanned version of said FRGCv2 subset. Bona fide and morphed face images are automatically generated employing four different morphing algorithms. The proposed PRNU-based morphing attack detector is shown to robustly distinguish bona fide and morphed face images achieving an average D-EER of 11.2% in the best configuration. In scenarios where image sources and morphing techniques are unknown, it is shown to significantly outperform other previously established morphing attack detectors. Finally, the limitations and potential of the approach are demonstrated on a dataset of printed and scanned bona fide and morphed face images.

Tackling In-Camera Downsizing for Reliable Camera ID Verification

The photo-response non-uniformity (PRNU) of an imaging sensor can be regarded as a biometric identifier unique to each camera. This modality is referred to as camera ID. The underlying process for estimating and matching camera IDs is now well established, and its robustness has been studied under a variety of processing. However, the effect of in-camera downsizing on camera ID verification has not yet been methodologically addressed. In this work, we investigate limitations imposed by built-in camera downsizing methods and tackle the question of how to obtain a camera ID so that attribution is possible with lower resolution media. For this purpose, we developed an application that gathers photos and videos at all supported resolutions by controlling camera settings. Analysis of media obtained from 21 smartphone and tablet cameras shows that downsizing of photos by a factor of 4 or higher suppresses PRNU pattern significantly. On the contrary, it is observed that source of unstabilized videos can be verified quite reliably at almost all resolutions. We combined our observations in a camera ID verification procedure considering downsized media.

Camera recognition with deep learning

In this paper, camera recognition with the use of deep learning technique is introduced. To identify the various cameras, their characteristic photo-response non-uniformity (PRNU) noise pattern was extracted. In forensic science, it is important, especially for child pornography cases, to link a photo or a set of photos to a specific camera. Deep learning is a sub-field of machine learning which trains the computer as a human brain to recognize similarities and differences by scanning it, in order to identify an object. The innovation of this research is the use of PRNU noise patterns and a deep learning technique in order to achieve camera identification. In this paper, AlexNet was modified producing an improved training procedure with high maximum accuracy of 80%–90%. DIGITS showed to have identified correctly six cameras out of 10 with a success rate higher than 75% in the database. However, many of the cameras were falsely identified indicating a fault occurring during the procedure. A possible explanation for this is that the PRNU signal is based on the quality of the sensor and the artefacts introduced during the production process of the camera. Some manufacturers may use the same or similar imaging sensors, which could result in similar PRNU noise patterns. In an attempt to form a database which contained different cameras of the same model as different categories, the accuracy rate was low. This provided further proof of the limitations of this technique, since PRNU is stochastic in nature and should be able to distinguish between different cameras from the same brand. Therefore, this study showed that current convolutional neural networks (CNNs) cannot achieve individualization with PRNU patterns. Nevertheless, the paper provided material for further research.

Source camera identification using Photo Response Non-Uniformity on WhatsApp

The Photo Response Non-Uniformity pattern can be a method for identification for an individual camera and is often present in digital footage. Therefore, the PRNU-pattern is also called the fingerprint of the camera. This pattern can be extracted and used to identify the source camera with a high likelihood ratio. This can be useful in cases such as child abuse or child pornography. In this research a 2nd order (FSTV) based method is used to extract the PRNU-patterns from videos of ten different mobile phone cameras. By calculating the Peak to Correlation Energy the PRNU-patterns of the natural videos are compared to the PRNU-patterns of the reference flat field videos of each camera to identify the source camera. This has been done for the original videos and the transmitted videos by WhatsApp for Android and IOS to determine if source camera identification by using PRNU is possible when videos are transmitted by WhatsApp. Also the PRNU-patterns of the natural videos are compared to each other to determine the possibility to find out if videos originate from the same source. With most cameras tested the method provides a high likelihood ratio, however for each case a validation of the method is necessary with reference cameras of the same model and type if used in casework. With videos transmitted by the IOS version of Whatsapp the source camera identification was not possible anymore.

Compressed Fingerprint Matching and Camera Identification via Random Projections

Sensor imperfections in the form of photoresponse nonuniformity (PRNU) patterns are a well-established fingerprinting technique to link pictures to the camera sensors that acquired them. The noise-like characteristics of the PRNU pattern make it a difficult object to compress, thus hindering many interesting applications that would require storage of a large number of fingerprints or transmission over a bandlimited channel for real-time camera matching. In this paper, we propose to use real-valued or binary random projections to effectively compress the fingerprints at a small cost in terms of matching accuracy. The performance of randomly projected fingerprints is analyzed from a theoretical standpoint and experimentally verified on databases of real photographs. Practical issues concerning the complexity of implementing random projections are also addressed using circulant matrices.

The effects of switching the camera module from BlackBerry Curve 9360 devices

A Photo-Response Non-Uniformity (PRNU) pattern is the ‘fingerprint’ of a digital camera, which is left in the images acquired with the camera. It can be used to identify the source of suspect images. For a case a BlackBerry phone with a camera was obtained and the question was if this phone was the source of certain suspect images. The phone was inaccessible, but the camera module was available for analysis. In this paper it is investigated if switching the camera module from one phone to another has an effect on the PRNU pattern. Results show that there is no significant difference between photos taken with the camera module still in the original phone and photos taken with the same camera module placed in a different phone. It is therefore possible to switch the camera module and use it for camera source identification, in cases where the original device is inaccessible.

Source Camera Identification for Heavily JPEG Compressed Low Resolution Still Images*

In this research, we examined whether fixed pattern noise or more specifically Photo Response Non-Uniformity (PRNU) can be used to identify the source camera of heavily JPEG compressed digital photographs of resolution 640 x 480 pixels. We extracted PRNU patterns from both reference and questioned images using a two-dimensional Gaussian filter and compared these patterns by calculating the correlation coefficient between them. Both the closed and open-set problems were addressed, leading the problems in the closed set to high accuracies for 83% for single images and 100% for around 20 simultaneously identified questioned images. The correct source camera was chosen from a set of 38 cameras of four different types. For the open-set problem, decision levels were obtained for several numbers of simultaneously identified questioned images. The corresponding false rejection rates were unsatisfactory for single images but improved for simultaneous identification of multiple images.