Combining PRNU and noiseprint for robust and efficient device source identification

Cozzolino, Davide; Marra, Francesco; Gragnaniello, Diego; Poggi, Gianfranco; Verdoliva, Luisa

doi:10.1186/s13635-020-0101-7

Cited by 49 publications

(8 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Source camera clustering based on PRNU, for a criminal case dataset, is presented in [ 35 ]. Device identification based on camera sensor is proposed in [ 17 ], where noise print information is used to support PRNU for camera source identification. Three classification methods were used: SVM, Likelihood ratio test, and Fisher’s linear discriminant.…”

Section: Related Workmentioning

confidence: 99%

Smartphone Camera Identification from Low-Mid Frequency DCT Coefficients of Dark Images

Berdich

Groza

2022

Entropy

View full text Add to dashboard Cite

Camera sensor identification can have numerous forensics and authentication applications. In this work, we follow an identification methodology for smartphone camera sensors using properties of the Dark Signal Nonuniformity (DSNU) in the collected images. This requires taking dark pictures, which the users can easily do by keeping the phone against their palm, and has already been proposed by various works. From such pictures, we extract low and mid frequency AC coefficients from the DCT (Discrete Cosine Transform) and classify the data with the help of machine learning techniques. Traditional algorithms such as KNN (K-Nearest Neighbor) give reasonable results in the classification, but we obtain the best results with a wide neural network, which, despite its simplicity, surpassed even a more complex network architecture that we tried. Our analysis showed that the blue channel provided the best separation, which is in contrast to previous works that have recommended the green channel for its higher encoding power.

show abstract

Section: Related Workmentioning

confidence: 99%

Smartphone Camera Identification from Low-Mid Frequency DCT Coefficients of Dark Images

Berdich

Groza

2022

Entropy

View full text Add to dashboard Cite

show abstract

“…e camera feature-based forensics mainly adopts inconsistent characteristics such as the distribution or relationship of camera color [75,79], camera pattern noise [80,81], photo response non-uniformity (PRNU) [82][83][84], and color filter array (CFA) [85,86] to realize image forensics.…”

Section: Image Generation Feature-based Forensicsmentioning

confidence: 99%

Overview of Image Inpainting and Forensic Technology

Liu

Bukhari

2022

Security and Communication Networks

View full text Add to dashboard Cite

At present, digital image is the most significant information carrier on the Internet, which provides convenience for people to exchange information. With the progress of image processing theory, digital image inpainting technology has also developed rapidly. It not only enhances the image expression but also provides tools for image forgery. Misusing the tools may trigger a crisis of trust in the image. To eliminate the negative impact of image forgery, many scholars have conducted in-depth research on it and proposed a series of detection methods called image forensics for institutions or communities to identify such forged images. Current forensic technology still has many limitations, such as relying on specific features and data distribution which makes forensic technology lag behind image inpainting technology. So far, academia still lacks a unified understanding of image forgery and detection technology, and the research architecture of inpainting and forensic technology is not clear. This article reviews the development of image inpainting and forensic technology and systematically summarizes and scientifically classifies the current research work. Finally, the forensic technology is summarized and the future research direction is prospected to guide subsequent scholars to further promote the progress of image forensic technology.

show abstract

“…Cozzolino et al [29] introduce a siamese method based on [2] to estimate camera-based fingerprints (called Noiseprints) for video with no need of prior knowledge on the specific manipulation or any form of fine-tuning. Another work [30] from the same research group combines the PRNU and Noiseprint to boost the performance of PRNUbased analyses based on only a few images. In some works [8,31,32] video file containers have been considered for the source identification of videos without a prior training phase.…”

Section: Forensic Analysismentioning

confidence: 99%

Identification of Social-Media Platform of Videos through the Use of Shared Features

Maiano

Amerini

Celsi³

et al. 2021

J. Imaging

View full text Add to dashboard Cite

Videos have become a powerful tool for spreading illegal content such as military propaganda, revenge porn, or bullying through social networks. To counter these illegal activities, it has become essential to try new methods to verify the origin of videos from these platforms. However, collecting datasets large enough to train neural networks for this task has become difficult because of the privacy regulations that have been enacted in recent years. To mitigate this limitation, in this work we propose two different solutions based on transfer learning and multitask learning to determine whether a video has been uploaded from or downloaded to a specific social platform through the use of shared features with images trained on the same task. By transferring features from the shallowest to the deepest levels of the network from the image task to videos, we measure the amount of information shared between these two tasks. Then, we introduce a model based on multitask learning, which learns from both tasks simultaneously. The promising experimental results show, in particular, the effectiveness of the multitask approach. According to our knowledge, this is the first work that addresses the problem of social media platform identification of videos through the use of shared features.

show abstract

Combining PRNU and noiseprint for robust and efficient device source identification

Cited by 49 publications

References 26 publications

Smartphone Camera Identification from Low-Mid Frequency DCT Coefficients of Dark Images

Smartphone Camera Identification from Low-Mid Frequency DCT Coefficients of Dark Images

Overview of Image Inpainting and Forensic Technology

Identification of Social-Media Platform of Videos through the Use of Shared Features

Contact Info

Product

Resources

About