Aug 2026· International Journal of Research Publication and Reviews· 0 citations
TL;DR
A fusion-based lightweight deep learning framework for copy-move image forgery detection and localization that offers an efficient and practical solution for digital image authentication and is applicable to digital forensics, journalism, law enforcement, cyber security, and multimedia content verification.
Abstract
The increasing accessibility of digital image editing tools has made image manipulation easier than ever before, creating significant challenges in verifying the authenticity of digital visual content. Copy-move forgery is one of the most widely used image manipulation techniques because duplicated regions originate from the same image, making manual detection difficult. Existing forgery detection approaches often rely on handcrafted feature extraction or a single deep learning architecture, which may exhibit reduced performance when images undergo transformations such as rotation, scaling, compression, illumination variation, and noise addition. This paper presents a fusion-based lightweight deep learning framework for copy-move image forgery detection and localization. The proposed system integrates three efficient convolutional neural network architectures, namely SqueezeNet, MobileNetV2, and ShuffleNet, to extract complementary deep image features. Instead of depending on a single classifier, the extracted features are fused and classified using a Support Vector Machine (SVM), resulting in improved prediction reliability and reduced false detection. To provide visual evidence of manipulation, the framework incorporates a Discrete Cosine Transform (DCT)-based localization module that identifies duplicated image regions and highlights forged areas. The proposed model was implemented using Python with TensorFlow, OpenCV, Scikit-learn, NumPy, and Tkinter. Experimental evaluation was conducted on the MICC-F220 benchmark dataset. The fusion strategy achieved an overall detection accuracy of 90.91%, outperforming individual lightweight convolutional neural network models while maintaining low computational complexity. The proposed framework offers an efficient and practical solution for digital image authentication and is applicable to digital forensics, journalism, law enforcement, cyber security, and multimedia content verification.
In the era of advanced digital image generation techniques such as copy-move forgery, the NDV is having increasing concerns regarding image authenticity particularly with the spread of digital images through social media, media and courts. This type of forgery, where a part of an image is replicated and pasted on the same image, to cover objects or materials poses severe challenges to the traditional approaches due to its inherent compactness in color, texture and noise. This research primarily focuses on the problem that deep learning models currently do not have sufficient interpretability and performance for the reliable detection and explanation of such forgeries under different image manipulations such as scaling, rotation and compression. This study proposes an explainable deep learning model which is capable of detecting copy-move forgery with high accuracy and is explainable as well from both feature and visual insights. The proposed technique has a hybrid CNN with both attention and feature matching modules, and an explainability module with Gradient-weighted Class Activation Mapping (Grad-CAM). The suggested framework has been evaluated on the benchmark datasets CoMoFoD and CASIA where accuracy, precision, recall and F1-scores were found to be 98.3, 97.6, 98.1 and 97.8 respectively which is higher than the other five state-of-the-art methods by 2-4 points. The approach was also able to withstand geometric transformations and compression and explainable by heatmap imaging. In conclusion, the proposed framework improves the detection and the interpretability of the deep learning-based copy-move forgery detection system, therefore it can be used in the domain of practical forensic research.
Shaheena K. V., D. S· International Conference on...· 0 citations
Increasing availability of advanced AI-based image creation and editing tools has significantly increased the prevalence of digital image forgery, creating challenges for media authenticity, digital forensics, and information security. This paper presents an intelligent image forgery detection framework that combines multi-quality Error Level Analysis (ELA) with an optimized Convolutional Neural Network (CNN) and heat map-based manipulation localization. The proposed approach utilizes ELA preprocessing at multiple JPEG quality levels to amplify compression inconsistencies associated with image tampering, this is followed by a lightweight CNN architecture trained using advanced optimization techniques, including warmup and cosine annealing learning-rate scheduling, stochastic gradient descent with momentum, and class imbalance handling. To improve interpretability, a heat map generation module is introduced to localize suspicious image regions and provide visual evidence supporting classification decisions which provides interpretability and explainability. The framework was evaluated using the CASIA 2.0 image forgery dataset containing a total of 12,614 images, 7,491 authentic and 5,123 tampered images. Experimental results demonstrate an overall classification accuracy of 94%, outperforming a baseline ELA-CNN implementation while maintaining computational efficiency. The proposed localization mechanism enhances explainability by highlighting potential manipulation regions. A web-based interface and RESTful API were developed to support practical deployment and integration into digital forensic workflows. The results indicate that the proposed framework provides an effective and interpretable smart solution for automated image forgery detection and localization.
Ahmad AlMunayyer, Abdallah Banat, Mohammad AlHayajneh et al.· IEEE Jordan Conference on Ap...· 0 citations
Artificial intelligence has significantly improved digital image editing capabilities, making it increasingly difficult to distinguish authentic images from manipulated ones [5, 7]. This paper proposes a Vision Transformer (ViT)-based framework for digital image forgery detection and localization by leveraging global contextual feature learning [4]. Unlike conventional Convolu-tional Neural Networks (CNNs), Vision Transformers capture long-range dependencies through self-attention mechanisms, enabling more effective identification of manipulated regions [4, 9]. The proposed framework performs image preprocessing, patch extraction, positional encod-ing, transformer-based feature learning, binary classification, and forgery localization. The model is evaluated using publicly available benchmark datasets, including CASIA V2, Co-MoFoD, and FaceForensics++ [20, 48], and its performance is assessed using Accuracy, Pre-cision, Recall, F1-score, Area Under Curve (AUC), Intersection over Union (IoU), and Pixel Accuracy [17, 49]. Experimental results demonstrate that the proposed Vision Transformer framework outperforms conventional CNN-based methods in terms of detection accuracy and localization precision [16, 19]. The proposed approach provides a robust and scalable solution for modern digital image forensics [15] and can be extended to hybrid transformer architectures and video forgery detection in future work.
G. Mary Pushpa, Dr. K. Sravan Adbhilash· International Journal of Lat...· 0 citations
ABSTRACT:
Digital images are widely used in social media, journalism, legal evidence, and scientific applications. The availability of advanced image editing tools has increased the risk of digital image forgery, which can lead to misinformation, privacy issues, security threats, and legal complications. Therefore, detecting image forgery accurately has become an important challenge in digital forensics. This paper presents an Image Forgery Detection System using Machine Learning techniques. The proposed system combines Error Level Analysis (ELA) with a Convolutional Neural Network (CNN) to identify manipulated images. ELA is used to highlight inconsistent compression patterns and possible tampered regions, while the CNN learns important visual features for classification. The system classifies uploaded images as either authentic or forged and provides a confidence score along with the prediction. A Python-based Flask backend is used for image processing and model integration, while HTML and CSS provide a simple and user-friendly interface. The results reported in the project demonstrate more than 90% accuracy on the test dataset. The proposed system reduces manual inspection and provides a faster approach for image authenticity verification. The system can be useful in digital forensics, media verification, security, and other applications where image authenticity is important. Future enhancements can include multi-class forgery classification, real-time detection, and cloud-based deployment.Top of Form
M. Tharani, G. Jayanth· International Scientific Jou...· 0 citations
Digital images have become the primary source of information exchange on digital platforms, making digital image forgery detection important and crucial in the field of digital forensics. Among various illegal manipulation techniques, copy-move and splicing manipulation are the two most common types of forgery. Copy-move forgery involves duplicating the areas within the same image, whereas splicing forgery involves taking out areas or objects from one or more images and combining them together to produce a spliced image. Detecting various forgeries in an image has become a crucial requirement due to the presence of subtle artifacts and minimal inconsistencies during the forgery process. Hence, a multi-forgery detection framework based on Groupwise-Spatial Attention–based Dense Neural Network (GSA-DNN) is proposed to detect copy-move and splicing forgery. Error level analysis (ELA) technique is used to detect image manipulation by figuring out the compression differences between authentic and forged images, followed by a generic standardizing procedure for all images based on dimensions and pixel values so as to improve final model convergence. InceptionNet V3 is employed to extract the multiscale features through convolutional kernels, which enables the learning of global and local feature representation. The proposed GSA-DNN method obtains high detection accuracy of 99.58%, 97.77%, and 94.59% on benchmark datasets such as CASIA V2, CoMoFoD, and Multiple Image Splicing Datasets compared to existing methods such as the Siamese neural network.
Received: 11 May 2026 | Revised: 28 July 2026 | Accepted: 14 August 2026
Conflicts of Interest
The authors declare that they have no conflicts of interest to this work.
Data Availability Statement
The data that support the findings of this study are openly available in CASIA V2 at https://www.mdpi.com/2076-3417/14/13/5545; in CoMoFoD at https://www.kaggle.com/datasets/tusharchauhan1898/comofod; in MISD at https://zenodo.org/records/5525829 and in CG-1050 at https://www.kaggle.com/datasets/saurabhshahane/cg1050.
Author Contribution Statement
S. Aruna: Conceptualization, Methodology, Formal analysis, Resources, Data curation, Writing – original draft. Surabhi Narayan: Software, Validation, Investigation, Writing – review & editing, Visualization, Supervision, Project administration.
S. Aruna, Surabhi Narayan· Artificial Intelligence and...· 0 citations
Experimental results demonstrate that the proposed approach effectively identifies deepfake images with high accuracy, making it suitable for applications in digital forensics, media verification, and cybersecurity.
J. Kollu, Mortha Pavan, Putta Vardhan et al.· International Journal of Inn...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
