Deep Learning-Based Digital Image Forgery Detection System

Emad Ul Haq Qazi,Abdulrazaq Almorjan,Tanveer Zia

doi:10.3390/app12062851

Emad Ul Haq Qazi, Abdulrazaq Almorjan + Show 1 more

Open Access

https://doi.org/10.3390/app12062851

Copy DOI

Journal: Applied sciences	Publication Date: Mar 10, 2022
Citations: 19	License type: CC BY 4.0

Affiliation: Naif Arab University for Security Sciences

Abstract

The advancements of technology in every aspect of the current age are leading to the misuse of data. Researchers, therefore, face the challenging task of identifying these manipulated forms of data and distinguishing the real data from the manipulated. Splicing is one of the most common techniques used for digital image tampering; a selected area copied from the same or another image is pasted in an image. Image forgery detection is considered a reliable way to verify the authenticity of digital images. In this study, we proposed an approach based on the state-of-the-art deep learning architecture of ResNet50v2. The proposed model takes image batches as input and utilizes the weights of a YOLO convolutional neural network (CNN) by using the architecture of ResNet50v2. In this study, we used the CASIA_v1 and CASIA_v2 benchmark datasets, which contain two distinct categories, original and forgery, to detect image splicing. We used 80% of the data for the training and the remaining 20% for testing purposes. We also performed a comparative analysis between existing approaches and our proposed system. We evaluated the performance of our technique with the CASIA_v1 and CASIA_v2 datasets. Since the CASIA_v2 dataset is more comprehensive compared to the CASIA_v1 dataset, we obtained 99.3% accuracy for the fine-tuned model using transfer learning and 81% accuracy without transfer learning with the CASIA_v2 dataset. The results show the superiority of the proposed system.

Highlights

Digital images have an important role in many fields such as in newspapers, digital forensics, scientific research, medicine, and so forth
We proposed an architecture using ResNet50v2 as our base model, and we used the YOLO convolutional neural network (CNN) weights for transfer learning
We present a deep learning-based architecture that uses a transfer learning technique to utilize the weights of a YOLO CNN

Summary

Introduction

Digital images have an important role in many fields such as in newspapers, digital forensics, scientific research, medicine, and so forth. Any pre-embedded information, such as a watermark embedded for the detection of image forgery, cannot be relied upon. This approach is known as the blind approach because there is no additional information for image forgery detection This approach is based on features that are extracted directly from the images. Some of the examples of CNN-based feature extractions are deep features utilized for image quality assessment [6], skin lesion classification [7], or person re-identification [8]. These extracted features are adapted into the inherent structural patterns of the data. We present an architecture based on the ResNet50v2 architecture that employs the use of transfer learning for the detection of tampered images, spliced images.

Literature Review

Proposed System Architecture

CASIA ITDE v2

Methodology

Future Work

Findings

Conclusions