FIDF: A Hypergraph Convolutional Framework for Detection of Social Media Image Forgery in Compression Artefacts

Md. Mehedi Rahman Rana; Md. Anisur Rahman; Kamrul Hasan Talukder; Rohul Amin

doi:10.53808/KUS.2026.23.01.1664-se

Authors

Md. Mehedi Rahman Rana Computer Science and Engineering Discipline, Science Engineering and Technology School, Khulna University, Khulna – 9208, Bangladesh
Md. Anisur Rahman Computer Science and Engineering Discipline, Science Engineering and Technology School, Khulna University, Khulna – 9208, Bangladesh
Kamrul Hasan Talukder Computer Science and Engineering Discipline, Science Engineering and Technology School, Khulna University, Khulna – 9208, Bangladesh
Rohul Amin Department of Computer Science and Engineering, Bangladesh Army University of Science and Technology, Siromoni, 9402, Khulna, Bangladesh

DOI:

https://doi.org/10.53808/KUS.2026.23.01.1664-se

Keywords:

Image Forgery Detection, Hypergraph Neural Networks, Deep Learning-Based Forgery, Social Media Image Forgery, Frequency Domain Analysis

Abstract

The rapid growth of manipulated images on social media has posed the challenging problem of robust image forgery detection. Existing methods either rely on content-driven deep models or single-view forgery cues, which limit their generalization ability under different types of manipulations and imaging conditions. In this paper, we propose FIDF (Fake Image Detection Framework), a novel multi-view discrepancy-gated hypergraph framework for detecting social media image forgeries. This motivates the proposed method based on the key observation that forged images introduce inconsistencies across multiple physically grounded representations. To take advantage of this, FIDF combines three complementary streams of features: a semantic RGB encoder, a learnable spectral representation based on a learnable DCT basis and a noise residual stream initialized from Spatial Rich Model (SRM) filters. These streams are fused by a Discrepancy-Gated Cross-Attention (DGCA) mechanism, which explicitly boosts the inter-view disagreement signals indicative of tampering. The k-nearest-neighbor hypergraph convolution module captures the non-local relations between spatially disjoint, but semantically similar regions and enables effective reasoning over complex manipulation patterns. We also present FIDD-13000 (Fake Image Detection Dataset-13000), a large-scale benchmark that reflects realistic social media forgery scenarios with different manipulation types and compression artefacts. Extensive experiments on FIDD-13000 and four public benchmarks (CASIA v1, CASIA 2.0, Columbia and MICC-F2000) demonstrate that FIDF outperforms the state-of-the-art methods consistently, with superior accuracy, F1-score and AUC under challenging conditions. Additional ablation studies further confirm the roles of multi-view streams, DGCA fusion, and hypergraph reasoning modules.

Downloads

Download data is not yet available.

References

Ali, S. S., Ganapathi, I. I., Vu, N.-S., Ali, S. D., Saxena, N., & Werghi, N. (2022). Image forgery detection using deep learning by recompressing images. Electronics, 11, 403, doi.org/10.3390/electronics11030403

Amerini, I., Ballan, L., Caldelli, R., Del Bimbo, A., & Serra, G. (2011). A SIFT-based forensic method for copy-move attack detection and transformation recovery. IEEE Transactions on Information Forensics and Security, 6, 1099–1110, 10.1109/TIFS.2011.2129512

Asghar, K., Sun, X., Rosin, P. L., Saddique, M., Hussain, M., & Habib, Z. (2019). Edge–texture feature-based image forgery detection with cross-dataset evaluation. Machine Vision and Applications, 30, 1243–1262, doi.org/10.1007/s00138-019-01048-2

Barni, M., Bondi, L., Bonettini, N., Bestagini, P., Costanzo, A., Maggini, M., Tondi, B., & Tubaro, S. (2017). Aligned and non-aligned double JPEG detection using convolutional neural networks. Journal of Visual Communication and Image Representation, 49, 153–163, doi.org/10.1016/j.jvcir.2017.09.003

Bayar, B., & Stamm, M. C. (2016). A deep learning approach to universal image manipulation detection using a new convolutional layer. In Proceedings of the 4th ACM Workshop on Information Hiding and Multimedia Security (pp. 5–10), doi.org/10.1145/2909827.2930786

Bayar, B., & Stamm, M. C. (2018). Constrained convolutional neural networks: A new approach towards general purpose image manipulation detection. IEEE Transactions on Information Forensics and Security, 13, 2691–2706, 10.1109/TIFS.2018.2825953

Bunk, J., Bappy, J. H., Mohammed, T. M., Nataraj, L., Flenner, A., Manjunath, B. S., Peterson, L. (2017). Detection and localization of image forgeries using resampling features and deep learning. In IEEE CVPR Workshops (pp. 1881–1889), 10.1109/CVPRW.2017.235

Chakraborty, S., Chatterjee, K., & Dey, P. (2022). Discovering tampered image in social media using ELA and deep learning. SN Computer Science, 3, 392, doi.org/10.1007/s42979-022-01311-w

Chen, G., Du, C., Yu, Y., Hu, H., Duan, H., & Zhu, H. (2025). A deepfake image detection method based on a multi-graph attention network. Electronics, 14, 482, doi.org/10.3390/electronics14030482

Dell'Anna, S., Montibeller, A., & Boato, G. (2025). TrueFake: A Real World Case Dataset of Last Generation Fake Images also Shared on Social Networks. " 2025 International Joint Conference on Neural Networks (IJCNN), Rome, Italy, 2025, pp. 1-8, 10.1109/IJCNN64981.2025.11228911

Dolhansky, B., Bitton, J., Pflaum, B., Lu, J., Howes, R., Wang, M., & Ferrer, C. C. (2020). The deepfake detection challenge (DFDC) dataset. arXiv preprint arXiv:2006.07397, doi.org/10.48550/arXiv.2006.07397

Dong, J., Wang, W., & Tan, T. (2013). CASIA image tampering detection evaluation database. In 2013 IEEE China Summit and International Conference on Signal and Information Processing (pp. 422–426), 10.1109/ChinaSIP.2013.6 625374

Farid, H. (2019). Image forensics. Annual Review of Vision Science, 5, 549–573.

Fridrich, J., & Kodovský, J. (2012). Rich models for steganalysis of digital images. IEEE Transactions on Information Forensics and Security, 7, 868–882, 10.1109/TIFS.2012.2190402

Gardella, M., Musé, P., Morel, J.-M., & Colom, M. (2021). Forgery detection in digital images by multi-scale noise estimation. Journal of Imaging, 7, 119, doi.org/10.3390/jimaging7070119

Gorle, R., & Guttavelli, A. (2025). Enhanced Image Tampering Detection using Error Level Analysis and CNN. Engineering, Technology & Applied Science Research, 15, 19683–19689, doi.org/10.48084/etasr.9593

Heller, S., Rossetto, L., & Schuldt, H. (2018). The PS-Battles dataset — an image collection for image manipulation detection. arXiv preprint arXiv:1804.04866, doi.org/10.48550/arXiv.1804.04866

Hsu, Y.-F., & Chang, S.-F. (2006). Detecting image splicing using geometry invariants and camera characteristics consistency. In IEEE International Conference on Multimedia and Expo (ICME) (pp. 549–552), 10.1109/ICME.2006.262447

Jiang, L., Li, R., Wu, W., Qian, C., & Loy, C. C. (2020). DeeperForensics-1.0: A large-scale dataset for real-world face forgery detection. arXiv. https://doi.org/10.48550/arXiv.2001.03024

Kaur, S., Sinha, N., Jain, P., Koli, S., Sharma, A., & Lathwal, A. (2024). Enhanced image forgery detection using a hybrid approach: Integration of ELA, CNN, and XGBoost. International Journal of Performability Engineering, 20, 367–378, 10.23940/ijpe.24.06.p4.367378

Khoo, B., Phan, R. C.-W., & Lim, C.-H. (2022). Deepfake attribution: On the source identification of artificially generated images. Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery, 12, e1438, doi.org/10.1002/widm.1438

Kwan, C., & Larkin, J. (2019). Demosaicing of Bayer and CFA 2.0 patterns for low lighting images. Electronics, 8, 1444, doi.org/10.3390/electronics8121444

Li, W., Yuan, Y., & Yu, N. (2009). Passive detection of doctored JPEG image via block artifact grid extraction. Signal Processing, 89, 1821–1829, doi.org/10.1016/j.sigpro.2009.03.025

Li, Y., Yang, X., Sun, P., Qi, H., & Lyu, S. (2020). Celeb-DF: A large-scale challenging dataset for DeepFake forensics. arXiv. https://doi.org/10.48550/arXiv.1909.12962

Loshchilov, I., & Hutter, F. (2017). Decoupled weight decay regularization. arXiv preprint arXiv:1711.05101, doi.org/10.48550/arXiv.1711.05101

Mahfoudi, G., Tajini, B., Retraint, F., Morain-Nicolier, F., Dugelay, J. L., & Pic, M. (2019). DEFACTO: Image and face manipulation dataset. In 2019 27th European Signal Processing Conference (EUSIPCO) (pp. 1–5), 10.23919/EUSIPCO.2019.8903181

Mao, D. (2023). DeepfakeArt Challenge dataset. Kaggle. https://www.kaggle.com/datasets/danielmao2019/deepfakeart

Mareen, H., De Neve, L., Lambert, P., & Van Wallendael, G. (2023). Harmonizing image forgery detection & localization: Fusion of complementary approaches. Journal of Imaging, 10, 4, doi.org/10.3390/jimaging10010004

Ng, T.-T., Hsu, J., & Chang, S.-F. (2009). Columbia image splicing detection evaluation dataset. DVMM Lab, Columbia University, ee.columbia.edu/ln/dvmm/downloads/AuthSplicedDataSet/dlform.html

Qazi, E. U. H., Zia, T., & Almorjan, A. (2022). Deep learning-based digital image forgery detection system. Applied Sciences, 12, 2851, doi.org/10.3390/app12062851

Rana, M. M. R., Hasnat, A., & Rahaman, G. M. A. (2022). SMIFD-1000: Social media image forgery detection database. Forensic Science International: Digital Investigation, 41, 301392, doi.org/10.1016/j.fsidi.2022.301392

Rana, M. M. R., Rahman, M. A., & Talukder, K. H. (2023). FIDD-1500: Fake Image Detection Dataset. In 2023 26th International Conference on Computer and Information Technology (ICCIT) (pp. 1–6), 10.1109/ICCIT6045 9.2023.10441386

Rossler, A., Cozzolino, D., Verdoliva, L., Riess, C., Thies, J., & Nießner, M. (2019). FaceForensics++: Learning to detect manipulated facial images. arXiv. https://doi.org/10.48550/arXiv.1901.08971

Sharma, D. K., Singh, B., Agarwal, S., Garg, L., Kim, C., & Jung, K.-H. (2023). A survey of detection and mitigation for fake images on social media platforms. Applied Sciences, 13, 10980, doi.org/10.3390/app131910980

Shi, Z., Chen, H., & Zhang, D. (2025). Robustifying vision transformer for image forgery localization with multi-exit architectures. Pattern Recognition, 164, 111565, doi.org/10.1016/j.patcog.2025.111565

Singh, D., Singh, P., Jena, R., & Chakraborty, R. S. (2023). An image forensic technique based on JPEG ghosts. Multimedia Tools and Applications, 82, 14153–14169, doi.org/10.1007/s11042-022-13699-x

Singh, S., & Kumar, R. (2024). Image forgery detection: comprehensive review of digital forensics approaches. Journal of Computational Social Science, 7, 877–915, doi.org/10.1007/s42001-024-00265-8

Tralic, D., Zupancic, I., Grgic, S., & Grgic, M. (2013). CoMoFoD — New database for copy-move forgery detection. In Proceedings ELMAR-2013 (pp. 49–54), http://www.vcl.fer.hr/comofod.

Tyagi, S., & Yadav, D. (2023). MiniNet: A concise CNN for image forgery detection. Evolving Systems, 14, 545–556, doi.org/10.1007/s12530-022-09446-0

Verdoliva, L. (2020). Media forensics and deepfakes: An overview. IEEE Journal of Selected Topics in Signal Processing, 14, 910–932, 10.1109/JSTSP.2020.3002101

Wang, T., Liao, X., Chow, K. P., Lin, X., & Wang, Y. (2024). Deepfake detection: A comprehensive survey from the reliability perspective. ACM Computing Surveys, 57, 1–35, doi.org/10.1145/3699710

Zampoglou, M., Papadopoulos, S., & Kompatsiaris, Y. (2017). Large-scale evaluation of splicing localization algorithms for web images. Multimedia Tools and Applications, 76, 4801–4834, doi.org/10.1007/s11042-016-3795-2