Explainable Dual-Model Framework for AI-Generated Image and Image-Tampering Detection

We have described Trustify AI, a framework that checks whether a digital image is authentic and explains how it reached its answer. The framework combines two EfficientNet-B0 detectors with metadata inspection, a rule-based decision engine, and Grad-CAM explanations. The CIFAKE detector scored 89.56% on its benchmark and the CASIA detector 78.87% on a split free of leakage. The outside test, however, gave a very different picture. The AI-generation detector flagged 11 of 13 real photographs, which shows a large generalization gap. We take this as a strong reason to show several kinds of evidence and to report results openly, instead of relying on benchmark accuracy alone. The framework is a workable first step toward transparent authenticity checking, but it is not yet ready for high-stakes forensic use. Before that, it will need larger and more varied training data, calibrated thresholds, and independent validation.

[1]     I. Goodfellow et al., “Generative adversarial nets,” in Advances in Neural Information Processing Systems, vol. 27, 2014. [2]     J. Ho, A. Jain, and P. Abbeel, “Denoising diffusion probabilistic models,” in Advances in Neural Information Processing Systems, vol. 33, 2020, pp. 6840–6851. [3]     L. Verdoliva, “Media forensics and deepfakes: An overview,” IEEE J. Sel. Topics Signal Process., vol. 14, no. 5, pp. 910–932, 2020. [4]     S.-Y. Wang, O. Wang, R. Zhang, A. Owens, and A. A. Efros, “CNN-generated images are surprisingly easy to spot… for now,” in Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit., 2020, pp. 8695–8704. [5]     J. Dong, W. Wang, and T. Tan, “CASIA image tampering detection evaluation database,” in Proc. IEEE China Summit Int. Conf. Signal Inf. Process. (ChinaSIP), 2013, pp. 422–426. [6]     M. Tan and Q. V. Le, “EfficientNet: Rethinking model scaling for convolutional neural networks,” in Proc. 36th Int. Conf. Mach. Learn., 2019, pp. 6105–6114. [7]     J. J. Bird and A. Lotfi, “CIFAKE: Image classification and explainable identification of AI-generated synthetic images,” IEEE Access, vol. 12, pp. 15642–15650, 2024. [8]     R. R. Selvaraju, M. Cogswell, A. Das, R. Vedantam, D. Parikh, and D. Batra, “Grad-CAM: Visual explanations from deep networks via gradient-based localization,” in Proc. IEEE Int. Conf. Comput. Vis., 2017, pp. 618–626. [9]     J. Deng, W. Dong, R. Socher, L.-J. Li, K. Li, and L. Fei-Fei, “ImageNet: A large-scale hierarchical image database,” in Proc. IEEE Conf. Comput. Vis. Pattern Recognit., 2009, pp. 248–255.