Visible to the public Composition of Visual Feature Vector Pattern for Deep Learning in Image Forensics

TitleComposition of Visual Feature Vector Pattern for Deep Learning in Image Forensics
Publication TypeJournal Article
Year of Publication2020
AuthorsRhee, K. H.
JournalIEEE Access
KeywordsCNN, CNN deep learning, CNN hybrid model, CNN net layer, CNN structure, convolution, convolutional neural nets, Deep Learning, extraction method, feature extraction, feature vector, Filtering, Forgery, forgery image, forgery images, Human Behavior, image classification, Image forensics, image forensics detection scheme, inception module, information forensics, learning (artificial intelligence), Least squares approximations, least-squares solution, machine learning, median filter residual, median filtering detection, median filters, Metrics, pubcrawl, residual block, resilience, Resiliency, Scalability, suspicious image, visual feature vector pattern, visualization, visualized feature vector patterns, window size

In image forensics, to determine whether the image is impurely transformed, it extracts and examines the features included in the suspicious image. In general, the features extracted for the detection of forgery images are based on numerical values, so it is somewhat unreasonable to use in the CNN structure for image classification. In this paper, the extraction method of a feature vector is using a least-squares solution. Treat a suspicious image like a matrix and its solution to be coefficients as the feature vector. Get two solutions from two images of the original and its median filter residual (MFR). Subsequently, the two features were formed into a visualized pattern and then fed into CNN deep learning to classify the various transformed images. A new structure of the CNN net layer was also designed by hybrid with the inception module and the residual block to classify visualized feature vector patterns. The performance of the proposed image forensics detection (IFD) scheme was measured with the seven transformed types of image: average filtered (window size: 3 x 3), gaussian filtered (window size: 3 x 3), JPEG compressed (quality factor: 90, 70), median filtered (window size: 3 x 3, 5 x 5), and unaltered. The visualized patterns are fed into the image input layer of the designed CNN hybrid model. Throughout the experiment, the accuracy of median filtering detection was 98% over. Also, the area under the curve (AUC) by sensitivity (TP: true positive rate) and 1-specificity (FP: false positive rate) results of the proposed IFD scheme approached to `1' on the designed CNN hybrid model. Experimental results show high efficiency and performance to classify the various transformed images. Therefore, the grade evaluation of the proposed scheme is "Excellent (A)".

Citation Keyrhee_composition_2020