SavedAdd your journal

Journals / NATURENGS / 2026 / Volume: 7 Issue: 1

Transfer Learning-Based Detection of AI-Generated Image

Mehtap Ülker(BİTLİS EREN ÜNİVERSİTESİ)0000-0001-8680-8518
Journal
NATURENGS
Pages
—
Publication Date
2026-09-08
🏛️Bibliographic metadata and full text access for this article are provided via DergiPark.
DergiPark Record ↗
✨

AI Summary & Key Takeaways

AI Insight
⚡ TL;DR(Too Long; Didn't Read)

The results showed that Swin Transformer-Tiny achieved the best overall performance, reaching an F1-score of 88.64% and outperforming the other architectures.

📌 3 Key Takeaways

🎯
Objective & Hypothesis:Recent advances in artificial intelligence–based image generation have significantly increased the number of AI-generated images that closely resemble real visual content.
📊
Key Findings & Data:To address this problem, this study investigates the automatic classification of real and AI-generated flower images using fine-tuned transfer learning models.
💡
Conclusion & Impact:The results showed that Swin Transformer-Tiny achieved the best overall performance, reaching an F1-score of 88.64% and outperforming the other architectures.
🔬Design:Modeling & Simulation
📍Scope:Comprehensive Sample
⚙️Variables:Transfer Learning · Swin Transformer-Tiny · Synthetic flower image detection

Abstract

Recent advances in artificial intelligence–based image generation have significantly increased the number of AI-generated images that closely resemble real visual content. This rapid growth has introduced new challenges for visual authenticity verification, particularly in natural object domains characterized by highly complex textures and subtle structural variations. Flower imagery represents a particularly challenging benchmark setting, where fine-grained petal textures, delicate color transitions, and complex organic forms can be realistically mimicked by generative models. To address this problem, this study investigates the automatic classification of real and AI-generated flower images using fine-tuned transfer learning models. For this purpose, a novel flower image dataset containing both real and DALL·E-generated synthetic samples was constructed and used as a benchmark for comparative evaluation. Four widely used transfer learning architectures, EfficientNet-B2, EfficientNet-B3, ConvNeXt-Tiny, and Swin Transformer-Tiny, were fine-tuned and evaluated. The results showed that Swin Transformer-Tiny achieved the best overall performance, reaching an F1-score of 88.64% and outperforming the other architectures.

🔖

Cite This Article

Ülker, M. (2026). Transfer Learning-Based Detection of AI-Generated Image. NATURENGS, Volume: 7 Issue: 1. https://doi.org/10.46572/naturengs.1887938

📄 Full Text PDF Viewer

📄

PDF görüntüleyici hazırlanıyor...

References (14)

  1. Goodfellow, I. J., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S.,. & Bengio, Y. (2014). Generative adversarial nets. Advances in neural information processing systems, 27.
  2. Rombach, R., Blattmann, A., Lorenz, D., Esser, P., & Ommer, B. (2022). High-resolution image synthesis with latent diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition (pp. 10684-10695).
  3. Verdoliva, L. (2020). Media forensics and deepfakes: an overview. IEEE journal of selected topics in signal processing, 14(5), 910-932.
  4. Chesney, B., & Citron, D. (2019). Deep fakes: A looming challenge for privacy, democracy, and national security. Calif. L. Rev., 107, 1753.
  5. Demir, L.E., and Canbay, Y., (2025). Deepfake Image Detection with Transfer Learning Models. Bitlis Eren Üniversitesi Fen Bilimleri Dergisi, 14(1), 546–560.
  6. Wang, S.Y., Wang, O., Zhang, R., Owens, A., and Efros, A.A., (2020). CNN-generated images are surprisingly easy to spot. for now. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 8695–8704.
  7. Raj, S., Mathew, J., and Mondal, A., (2024). Generalized and robust model for GAN-generated image detection. Pattern Recognition Letters, 182, 104–110.
  8. Khan, S. B., Gupta, M., Gopinathan, B., Thyluru RamaKrishna, M., Saraee, M., Mashat, A., & Almusharraf, A. (2025). DeepFake Detection: Evaluating the Performance of EfficientNetV2‐B2 on Real vs. Fake Image Classification. IET Image Processing, 19(1).
  9. Tan, M., and Le, Q., (2019). EfficientNet: Rethinking model scaling for convolutional neural networks. Proceedings of the International Conference on Machine Learning, 6105–6114.
  10. Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., and Chen, L.C., (2018). MobileNetV2: Inverted residuals and linear bottlenecks. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, 4510–4520.
  11. Tan, M., and Le, Q., (2021). EfficientNetV2: Smaller models and faster training. Proceedings of the International Conference on Machine Learning, 10096–10106.
  12. Liu, Z., Mao, H., Wu, C. Y., Feichtenhofer, C., Darrell, T., & Xie, S. (2022). A convnet for the 2020s. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition (pp. 11976-11986).
  13. Goodfellow, I. J., Pouget-Abadie, J., Mirza, M., Xu, B., Warde-Farley, D., Ozair, S.,. & Bengio, Y. (2014). Generative adversarial nets. Advances in neural information processing systems, 27.
  14. Wang, S.Y., Wang, O., Zhang, R., Owens, A., and Efros, A.A., (2020). CNN-generated images are surprisingly easy to spot. for now. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 8695–8704.