Fourier-Based GAN Fingerprint Detection using ResNet50
Fuente:
arXiv
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866908606686822400 |
|---|---|
| author | Erukude, Sai Teja Marella, Viswa Chaitanya Veluru, Suhasnadh Reddy |
| author_facet | Erukude, Sai Teja Marella, Viswa Chaitanya Veluru, Suhasnadh Reddy |
| contents | The rapid rise of photorealistic images produced from Generative Adversarial Networks (GANs) poses a serious challenge for image forensics and industrial systems requiring reliable content authenticity. This paper uses frequency-domain analysis combined with deep learning to solve the problem of distinguishing StyleGAN-generated images from real ones. Specifically, a two-dimensional Discrete Fourier Transform (2D DFT) was applied to transform images into the Fourier domain, where subtle periodic artifacts become detectable. A ResNet50 neural network is trained on these transformed images to differentiate between real and synthetic ones. The experiments demonstrate that the frequency-domain model achieves a 92.8 percent and an AUC of 0.95, significantly outperforming the equivalent model trained on raw spatial-domain images. These results indicate that the GAN-generated images have unique frequency-domain signatures or "fingerprints". The method proposed highlights the industrial potential of combining signal processing techniques and deep learning to enhance digital forensics and strengthen the trustworthiness of industrial AI systems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_19840 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Fourier-Based GAN Fingerprint Detection using ResNet50 Erukude, Sai Teja Marella, Viswa Chaitanya Veluru, Suhasnadh Reddy Computer Vision and Pattern Recognition The rapid rise of photorealistic images produced from Generative Adversarial Networks (GANs) poses a serious challenge for image forensics and industrial systems requiring reliable content authenticity. This paper uses frequency-domain analysis combined with deep learning to solve the problem of distinguishing StyleGAN-generated images from real ones. Specifically, a two-dimensional Discrete Fourier Transform (2D DFT) was applied to transform images into the Fourier domain, where subtle periodic artifacts become detectable. A ResNet50 neural network is trained on these transformed images to differentiate between real and synthetic ones. The experiments demonstrate that the frequency-domain model achieves a 92.8 percent and an AUC of 0.95, significantly outperforming the equivalent model trained on raw spatial-domain images. These results indicate that the GAN-generated images have unique frequency-domain signatures or "fingerprints". The method proposed highlights the industrial potential of combining signal processing techniques and deep learning to enhance digital forensics and strengthen the trustworthiness of industrial AI systems. |
| title | Fourier-Based GAN Fingerprint Detection using ResNet50 |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2510.19840 |