Resampling Forgery Detection Using Deep Learning and A-Contrario Analysis

Abstract

The amount of digital imagery recorded has recently grown exponentially, and with the advancement of software, such as Photoshop or Gimp, it has become easier to manipulate images. However, most images on the internet have not been manipulated and any automated manipulation detection algorithm must carefully control the false alarm rate. In this paper we discuss a method to automatically detect local resampling using deep learning while controlling the false alarm rate using a-contrario analysis. The automated procedure consists of three primary steps. First, resampling features are calculated for image blocks. A deep learning classifier is then used to generate a heatmap that indicates if the image block has been resampled. We expect some of these blocks to be falsely identified as resampled. We use a-contrario hypothesis testing to both identify if the patterns of the manipulated blocks indicate if the image has been tampered with and to localize the manipulation. We demonstrate that this strategy is effective in indicating if an image has been manipulated and localizing the manipulations. from the

Toward Reliable Models for Authenticating Multimedia Content: Detecting Resampling Artifacts With Bayesian Neural Networks

“…Several learning-based methods achieve similar goals, either by directly detecting resampling [22] or the resampling factor [23], or indirectly via camera-based image forgery localization [24] or the detection of image splicing [25]. However, learning-based methods are sensitive to mismatches between training and test data.…”

Improved Fully Convolutional Network for Digital Image Region Forgery Detection

“…Among the reported literature investigations concerning image tamper identification, some of them are focused mainly on using deep learning and neural network techniques, but they still suffer from some problems. For example, literature investigations [Bayar and Stamm (2016) ;Flenner, Peterson, Bunk et al (2018); Cui, McIntosh and Sun (2018)] have applied deep learning techniques for image tamper identification, which were found to have the ability to solve the single tampering problem, but they could not solve the problem of detecting image splicing behavior. Furthermore, some methods were proposed in the literature [Cozzolino and Verdoliva (2016); Bondi, Lameri, Güera et al (2017)] to solve the image splicing problem, but they were based on some certain assumptions and thus greatly reduced the general applicability of the algorithms.…”

Extraction of image resampling using correlation aware convolution neural networks for image tampering detection

“…Due to interpolation resampling introduces periodic correlation between the pixels. The CNN-based tampering detection methodologies shows good translational invariance to produce spatial maps across different segment of multimedia content, and certain artifacts are well-learned using resampling feature sets [27]; which can be utilized to locate tampered segments [28], [29]. From extensive, it can be seen resampling feature detection of hybrid attacks within copy-clones attacks is a challenging task.…”