Generalization in Deepfake Detection: : A Comparative Analysis of Frequency-Domain and CNN Approaches

Nathalia Farinha Rodrigues; Ademar Takeo Akabane; Vinicius Carbonezi de Souza; Luiz Antonio Buschetto Macarini

doi:10.22456/2175-2745.150900

Authors

Nathalia Farinha Rodrigues Centro de Pesquisa e Desenvolvimento em Telecomunicações (CPQD)
Ademar Takeo Akabane Pontifícia Universidade Católica de Campinas (PUC-Campinas) https://orcid.org/0000-0002-4930-182X
Vinicius Carbonezi de Souza Centro de Pesquisa e Desenvolvimento em Telecomunicações (CPQD) https://orcid.org/0009-0006-1820-2825
Luiz Antonio Buschetto Macarini Centro de Pesquisa e Desenvolvimento em Telecomunicações (CPQD)

DOI:

https://doi.org/10.22456/2175-2745.150900

Keywords:

deepfake detection, generalization, generative adversarial networks, convolutional neural networks, spectral analysis, cross-database evaluation

Abstract

This study investigates deepfake detection, a rapidly evolving class of synthetic media with major financial and ethical implications. We evaluate the generalization of two detection approaches: Convolutional Neural Network (CNN)-based models, represented by EfficientNetAutoAttB4, and frequency-domain models, exemplified by FreqNet. While EfficientNetAutoAttB4 achieves strong performance on in-domain data, it tends to overfit, whereas FreqNet captures spectral artifacts that improve robustness to unseen manipulations. Cross-database evaluation on the GANGen-Detection dataset shows that FreqNet consistently outperforms EfficientNet across Accuracy, AUC-ROC, and F1-score. These results highlight the importance of frequency-domain representations for building more generalizable deepfake detectors and expose the limitations of purely spatial CNNs when facing out-of-distribution data. Code and example images to reproduce the experiments are available in the associated repository https://github.com/NathFarinha/deepfake-detection-generalization-efficientnet-freqnet.git.

Downloads

Download data is not yet available.

References

[1] MIRSKY, Y.; LEE, W. The creation and detection of deepfakes: A survey. ACM Computing Surveys (CSUR), ACM New York, NY, USA, v. 54, n. 1, p. 1–41, 2021.

[2] MCAFEE. McAfee Threat Report 2025: Artificial Intelligence and Emerging Threats. [S.l.], 2025. Disponível em: https://www.mcafee.com/blogs/wp-content/uploads/2025/01/McAfee-State-of-the-Scamiverse-2025.pdf.

[3] TAN, M.; LE, Q. EfficientNet: Rethinking model scaling for convolutional neural networks. In: PMLR. International Conference on Machine Learning. [S.l.], 2019. p. 6105–6114.

[4] YAN, Z. et al. DeepfakeBench: A comprehensive benchmark of deepfake detection. 2023. ArXiv preprint arXiv:2307.01426. Disponível em: https://arxiv.org/abs/2307.01426.

[5] TAN, C. et al. Frequency-aware deepfake detection: Improving generalizability through frequency space domain learning. In: Proceedings of the AAAI Conference on Artificial Intelligence. [S.l.: s.n.], 2024. v. 38, n. 5, p. 5052–5060.

[6] BONETTINI, N. et al. Video face manipulation detection through ensemble of CNNs. In: 2020 25th International Conference on Pattern Recognition (ICPR). [S.l.: s.n.], 2021. p. 5012–5019.

[7] GUARNERA, L.; GIUDICE, O.; BATTIATO, S. Deepfake detection by analyzing convolutional traces. IEEE Access, IEEE, v. 8, p. 165085–165098, 2020.

[8] NGUYEN, H. H.; YAMAGISHI, J.; ECHIZEN, I. Multi-task learning for detecting and segmenting manipulated facial images and videos. In: IEEE. 2019 IEEE 10th International Conference on Biometrics Theory, Applications and Systems (BTAS). [S.l.], 2019. p. 1–8.

[9] DOLHANSKY, B. et al. The Deepfake Detection Challenge (DFDC) dataset. 2020. ArXiv preprint arXiv:2006.07397. Disponível em: https://arxiv.org/abs/2006.07397.

[10] DURALL, R.; KEUPER, M.; KEUPER, J. Watch your up-convolution: CNN based generative deep neural networks are failing to reproduce spectral distributions. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. [S.l.: s.n.], 2020. p. 7890–7899.

[11] FRANK, J. et al. Leveraging frequency analysis for deep fake image recognition. In: PMLR. International Conference on Machine Learning. [S.l.], 2020. p. 3247–3258.

[12] QIAN, Y. et al. Thinking in frequency: Face forgery detection by mining frequency-aware clues. In: SPRINGER. European Conference on Computer Vision. [S.l.], 2020. p. 86–103.

[13] TAN RENSHUAI TAO, H. L. Y. Z. C. GANGen-Detection: A Dataset generated by GANs for Generalizable deepfake Detection. [S.l.]: GitHub, 2024. https://github.com/chuangchuangtan/GANGen-Detection.

[14] HE, Z. et al. AttGAN: Facial attribute editing by only changing what you want. IEEE Transactions on Image Processing, IEEE, v. 28, n. 11, p. 5464–5478, 2019.

[15] BERTHELOT, D.; SCHUMM, T.; METZ, L. BEGAN: Boundary Equilibrium Generative Adversarial Networks. 2017. ArXiv preprint arXiv:1703.10717. Disponível em: https://arxiv.org/abs/1703.10717.

[16] BELLEMARE, M. G. et al. The Cramer distance as a solution to biased Wasserstein gradients. 2017. ArXiv preprint arXiv:1705.10743. Disponível em: https://arxiv.org/abs/1705.10743.

[17] HJELM, R. D. et al. Learning deep representations by mutual information estimation and maximization. 2018. ArXiv preprint arXiv:1808.06670. Disponível em: https://arxiv.org/abs/1808.06670.

[18] LI, C.-L. et al. MMD GAN: Towards deeper understanding of moment matching network. Advances in Neural Information Processing Systems, v. 30, 2017.

[19] MIYATO, T. et al. Spectral normalization for generative adversarial networks. 2018. ArXiv preprint arXiv:1802.05957. Disponível em: https://arxiv.org/abs/1802.05957.

[20] LIU, M. et al. STGAN: A unified selective transfer network for arbitrary image attribute editing. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. [S.l.: s.n.], 2019. p. 3673–3682.

[21] RÖSSLER, A. et al. FaceForensics++: Learning to detect manipulated facial images. In: Proceedings of the IEEE/CVF International Conference on Computer Vision. [S.l.: s.n.], 2019. p. 1–11.

[22] BISHOP, C. M. Pattern Recognition and Machine Learning. [S.l.]: Springer, 2006.

[23] SOKOLOVA, M.; LAPALME, G. A systematic analysis of performance measures for classification tasks. Information Processing & Management, Elsevier, v. 45, n. 4, p. 427–437, 2009.

[24] POWERS, D. M. Evaluation: From precision, recall and f-measure to roc, informedness, markedness and correlation. Journal of Machine Learning Technologies, v. 2, n. 1, p. 37–63, 2011.

[25] FAWCETT, T. An introduction to ROC analysis. Pattern Recognition Letters, Elsevier, v. 27, n. 8, p. 861–874, 2006.