Information Forensics: An Overview of the First Decade

Stamm, Matthew C.; Wu, Min; Liu, K. J. Ray

doi:10.1109/access.2013.2260814

Cited by 341 publications

(156 citation statements)

References 137 publications

Supporting

Mentioning

154

Contrasting

Unclassified

Order By: Relevance

“…In the following we describe the proposed algorithm to solve this problem 1 . We first explain how to perform the training step needed to learn the CNN and SVMs parameters.…”

Section: Cnn For Camera Model Identificationmentioning

confidence: 99%

“…The overall architecture is characterized by 340, 462 parameters, learned through Stochastic Gradient Descent on batches of 128 patches. Momentum is fixed to 0.9, weights decay is set 1 Code available at https: to 7.5 · 10 −3 while the learning rate is initialized to 0.015 and halves every 10 epochs. As trained CNN model M, we select the one that provides the smallest loss on validation patches within the first 50 training epochs.…”

Section: Cnn For Camera Model Identificationmentioning

confidence: 99%

“…As downloading, copying, forging and redistributing digital material is becoming easier over the years, some forms of regulation and authenticity verification have become of urgent necessity. To this purpose, the forensic community has developed a wide series of tools to reverse-engineer the history of multimedia objects [1], [2].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

First Steps Toward Camera Model Identification With Convolutional Neural Networks

Bondi

Baroffio

Güera

et al. 2017

IEEE Signal Process. Lett.

226

184

View full text Add to dashboard Cite

Abstract-Detecting the camera model used to shoot a picture enables to solve a wide series of forensic problems, from copyright infringement to ownership attribution. For this reason, the forensic community has developed a set of camera model identification algorithms that exploit characteristic traces left on acquired images by the processing pipelines specific of each camera model. In this paper, we investigate a novel approach to solve camera model identification problem. Specifically, we propose a data-driven algorithm based on convolutional neural networks, which learns features characterizing each camera model directly from the acquired pictures. Results on a wellknown dataset of 18 camera models show that: (i) the proposed method outperforms up-to-date state-of-the-art algorithms on classification of 64x64 color image patches; (ii) features learned by the proposed network generalize to camera models never used for training.

show abstract

“…In the following we describe the proposed algorithm to solve this problem 1 . We first explain how to perform the training step needed to learn the CNN and SVMs parameters.…”

Section: Cnn For Camera Model Identificationmentioning

confidence: 99%

Section: Cnn For Camera Model Identificationmentioning

confidence: 99%

See 1 more Smart Citation

First Steps Toward Camera Model Identification With Convolutional Neural Networks

Bondi

Baroffio

Güera

et al. 2017

IEEE Signal Process. Lett.

226

184

View full text Add to dashboard Cite

show abstract

“…99% of the photographs in their test set. Others have followed the same path (for an overview, see Stamm et al 2013), but, as far as we know, perfect computational solutions for differentiation are lacking. This is due to studies in computer science having to deal with a specific tension in the field: while some are especially interested in reliably distinguishing photographs from computer-generated images, others have the objective of creating photorealistic renderings for which this distinction cannot be made.…”

Section: Related Researchmentioning

confidence: 99%

Photorealistic computer-generated images are difficult to distinguish from digital photographs: a case study with professional photographers and photo-editors

Lehmuskallio

Häkkinen

Seppänen

2018

Visual Communication

View full text Add to dashboard Cite

Abstract:There are strict guidelines on photoediting in newsrooms, and serious professional repercussions if failing to adhere to these, while computer-generated imagery is increasingly used in other areas of visual communication. This paper presents empirical research on the ability of professional photographers and editors to distinguish photographs from photorealistic computer-generated images by looking at them on a screen. Our results show clearly that those studied (n= 20) are unable to distinguish these from another, suggesting that it is increasingly difficult to make this distinction, particularly since most viewers are not as experienced in photography as those studied. Interestingly, those studied continue to share a conventional understanding of photography, that is not in line with current developments in digital photography and digital image rendering. Based on our findings, we suggest the need for developing a particular visual literacy that understands the computational in digital photography, and grounds the use of digital photography among particular communities of practice. When seeing photographs on screens, journals, exhibitions, or newspapers, we might actually be looking at computer-generated simulations, and vice versa.

show abstract

“…In this regard, Ray Liu [9] account for the scheme which is highly accurate in unaltered or uncompressed images.…”

Section: ⅰ 서 론mentioning

confidence: 99%

Median Filtering Detection using Latent Growth Modeling

Rhee¹

2015

Journal of the Institute of Electronics and Information Enginee

View full text Add to dashboard Cite

In recent times, the median filtering (MF) detector as a forensic tool for the recovery of forgery images' processing history has concerned broad interest. For the classification of MF image, MF detector should be designed with smaller feature set and higher detection ratio. This paper presents a novel method for the detection of MF in altered images. It is transformed from BMP to several kinds of MF image by the median window size. The difference distribution values are computed according to the window sizes and then the values construct the feature set same as the MF window size. For the MF detector, the feature set transformed to the model specification which is computed using latent growth modeling (LGM). Through experiments, the test image is classified by the discriminant into two classes: the true positive (TP) and the false negative (FN). It confirms that the proposed algorithm is to be outstanding performance when the minimum distance average is 0.119 in the confusion of TP and FN for the effectivity of classification.

show abstract

Information Forensics: An Overview of the First Decade

Cited by 341 publications

References 137 publications

First Steps Toward Camera Model Identification With Convolutional Neural Networks

First Steps Toward Camera Model Identification With Convolutional Neural Networks

Photorealistic computer-generated images are difficult to distinguish from digital photographs: a case study with professional photographers and photo-editors

Median Filtering Detection using Latent Growth Modeling

Contact Info

Product

Resources

About