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

Abstract

<span>Detecting hybrid tampering attacks in an image is extremely difficult; especially when copy-clone tampered segments exhibit identical illumination and contrast level about genuine objects. The existing method fails to detect tampering when the image undergoes hybrid transformation such as scaling, rotation, compression, and also fails to detect under small-smooth tampering. The existing resampling feature extraction using the Deep learning techniques fails to obtain a good correlation among neighboring pixels in both horizontal and vertical directions. This work presents correlation aware convolution neural network (CA-CNN) for extracting resampling features for detecting hybrid tampering attacks. Here the image is resized for detecting tampering under a small-smooth region. The CA-CNN is composed of a three-layer horizontal, vertical, and correlated layer. The correlated layer is used for obtaining correlated resampling feature among horizontal sequence and vertical sequence. Then feature is aggregated and the descriptor is built. An experiment is conducted to evaluate the performance of the CA-CNN model over existing tampering detection methodologies considering the various datasets. From the result achieved it can be seen the CA-CNN is efficient considering various distortions and post-processing attacks such joint photographic expert group (JPEG) compression, and scaling. This model achieves much better accuracies, recall, precision, false positive rate (FPR), and F-measure compared existing methodologies.</span>

Tampering Detection and Segmentation Model for Multimedia Forensic

“…The work has used a total of three datasets; in this first, the image is resized to into the non-overlapping region of similar to the work presented in [17]; thus, induces certain artifacts. In [17] used space-filling curve for extracting correlation among both horizontal and vertical streams; the model achieved good, tampered region detection accuracies; however, with poor segmentation accuracies; especially under small-smooth hybrid attacks.…”

“…The work has used a total of three datasets; in this first, the image is resized to into the non-overlapping region of similar to the work presented in [17]; thus, induces certain artifacts. In [17] used space-filling curve for extracting correlation among both horizontal and vertical streams; the model achieved good, tampered region detection accuracies; however, with poor segmentation accuracies; especially under small-smooth hybrid attacks. In addressing the segmentation problem, this paper introduces an improved CNN model that with an additional layer to obtain a good correlated features-map for achieving improved tampered region detection and segmentation outcome.…”

“…Recently, DL models like auto-encoders [7] and Convolutional Neural Networks (CNN) [8], [9] have been employed for image tampering detection with good results. Existing tampering detection models are predominantly designed to detect only one kind of attack [10], [11]. In this way, one methodology probably won't excel in different sorts of tampering attacks.…”

“…The resampling attack generally occasionally allows correlation among pixels because of interpolation. The CNN-based [16], [17], image forgery identification model learns resampling features [18] very well using spatial maps produced through translation invariance of various regions of images [19], [20]. Thus, this research work aims to build an efficient resampling feature detection through CNN to detect hybrid attacks and achieve better-tampered region segmentation outcomes [21], [22].…”

Noise Invariant Convolution Neural Network for Segmentation of Multiple Sclerosis Lesions from Brain Magnetic Resonance Imaging

“…The inception module in a parallel manner uses 3 × 3, 5 × 5, 1 × 1, and max pooling filters. The improvised model achieves a better MS lesion segmentation outcome than the standard CNN model [32,33]. In [28] showed that the UNET architecture are very good in performing segmentation of tumor in Kidney using CT images.…”