Autors: Kotov, G. I., Nakov, O. N., Lazarova, M. K., Nakov, P. O.
Title: Hybrid CNN and Forensic Approach for Detecting AI-Generated Human Faces
Keywords: AI-generated faces, deepfakes, face recognition, Generative Adversarial Networks, image forensics

Abstract: The rapid development of deep generative models, particularly Generative Adversarial Networks (GANs) and diffusion models, has resulted in generation of synthetic images that closely mimic real human features. This growing realism has serious implications for security, digital identity verification, and misinformation. The paper presents an overview of the technological and scientific challenges related to distinguishing between real human face images and those generated by artificial intelligence. A hybrid detection framework is suggested that integrates deep CNN-based features with handcrafted forensic cues. The experimental results of training the suggested hybrid approach using WhichFaceIsReal dataset show that the hybrid model outperforms both CNN-only and forensic-only baselines achieving an accuracy of 94.8% and demonstrating improved precision, recall, and robustness.

References

  1. I. Goodfellow, et al., "Generative Adversarial Networks", Communications of the ACM, vol. 63, no. 11, pp. 139-144, 2020
  2. G. Hongchang, J. Pei and H. Huang, "ProGAN: Network Embedding via Proximity Generative Adversarial Network", Proc. of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, pp. 1308-131, 2019.
  3. I. Skorokhodov, S. Tulyakov and M. Elhoseiny, "StyleGAN-V: A Continuous Video Generator With the Price, Image Quality and Perks of StyleGAN2", Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 3626-3636, 2020.
  4. H. Qiu, Y. Jiang, H. Zhou, W. Wu and Z. Liu, "StyleFaceV: Face Video Generation via Decomposing and Recomposing Pretrained StyleGAN3", arXiv: 2208. 07862, 2022.
  5. A. Stöckl, "Evaluating a Synthetic Image Dataset Generated with Stable Diffusion", in X. Yang, R. Sherratt, N. Dey, A. Joshi (eds.) Proc. of Eighth International Congress on Information and Communication Technology, Lecture Notes in Networks and Systems, vol 693. Springer, Singapore, 2023.
  6. F.-A. Croitoru, V. Hondru, R. T. Ionescu and M. Shah, "Diffusion Models in Vision: A Survey", IEEE Transactions on PAMI, vol. 45, no. 9, pp. 10850-10869, 2023.
  7. G. Marcus, E. Davis and S. Aaronson, "A Very Preliminary Analysis of DALL-E 2", arXiv: 2204. 13807, 2022.
  8. H. Hu, et al., "Instruct-imagen: Image Generation With Multi-Modal Instruction", Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 4754-4763, 2024
  9. X. Wang, e al., "GAN-Generated Faces Detection: A Survey and New Perspectives", Frontiers in Artificial Intelligence and Applications, vol. 372, pp. 2533-2542, 2023.
  10. R. Ramanath, et al., "Spectral Spaces and Color Spaces", Color Research & Application, vol. 9, no. 1, pp. 29-37, 2003.
  11. S. Targ, D. Almeida and K. Lyman, "Resnet in Resnet: Generalizing Residual Architectures", arXiv: 1603. 08029, 2016.
  12. F. Chollet, "Xception: Deep Learning with Depthwise Separable Convolutions", Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, pp. pp. 1800-1807, 2017.
  13. M. Tan and Q. Le, "EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks, " Proc. of International Conference on Machine Learning, pp. 6105-6114, 2019.
  14. H. Dang, et al., "On the Detection of Digital Face Manipulation", Proc. of the IEEE/CVF Conf. on Computer Vision and Pattern Recognition, pp. 5781-5790, 2020.
  15. W. Liu, et al., "An Attention-Based Multiscale Transformer Network for Remote Sensing Image Change Detection", Journal of Photogr. and Remote Sensing, vol. 202, pp. 599-609, 2023.
  16. I. Camacho, "Initialization Methods of Convolutional Neural Networks For Detection of Image Manipulations", Université Grenoble Alpes, 2021.
  17. S.-Y. Wang et al., "CNN-Generated Images Are Surprisingly Easy to Spot for Now", Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pp. 8695-8704, 2020.
  18. G. K. Pandey and S. Srivastava, "ResNet-18 Comparative Analysis of Various Activation Functions for Image Classification", Proc. of IEEE International Conference on Inventive Computation Technologies, pp. 595-601, 2023.
  19. M. Groh, Z. Epstein, C. Firestone and R. Picard, "Deepfake Detection by Human Crowds, Machines, and Machine-informed Crowds", Proc. of the National Acad. of Sc., vol. 119, no. 1, 2022.
  20. D. Hajialigol, H. Liu and X. Wang, "XAI-CLASS: Explanation-Enhanced Text Classification with Extremely Weak Supervision", arXiv: 2311. 00189, 2023.
  21. A. Vishwa and F. K. Hussain, "A Blockchain Based Approach for Multimedia Privacy Protection and Provenance", IEEE Symp. Series on Comp. Intell., Bangalore, India, pp. 1941-1945, 2018.

Issue

60th International Scientific Conference on Information, Communication and Energy Systems and Technologies, ICEST 2025 - Proceedings, 2025, Albania, https://doi.org/10.1109/ICEST66328.2025.11098362

Вид: публикация в международен форум, публикация в реферирано издание, индексирана в Scopus