用小波变换+ResNet50识别生成图像指纹，准确率达95.1%。
Wavelet-based GAN Fingerprint Detection using ResNet50
- 输入图像经小波变换转为多分辨率特征，再交由ResNet50分类
- 达贝舒斯小波模型准确率95.1%，远超空间域模型的81.5%
- 适合关注深度伪造检测与图像取证的研究者
生成对抗网络（GAN）生成图像的识别已成为数字图像取证的重要挑战。本研究提出一种基于小波变换的检测方法，通过离散小波变换（DWT）预处理，结合ResNet50分类器区分风格化生成图像与真实图像。采用哈尔（Haar）与达贝舒斯（Daubechies）小波滤波器将输入图像转换为多分辨率表示，利用生成过程留下的细微痕迹进行判别。对比在空间域训练的相同ResNet50模型，哈尔与达贝舒斯预处理模型分别达到93.8%和95.1%的准确率，显著优于空间域模型的81.5%。达贝舒斯模型表现更优，表明引入多层频率描述可增强区分能力。结果表明，GAN生成图像在小波域存在独特伪影或‘指纹’。该方法验证了小波域分析在检测GAN图像中的有效性，并揭示了未来深度伪造检测系统的发展潜力。
原文摘要 · Abstract (English)
Identifying images generated by Generative Adversarial Networks (GANs) has become a significant challenge in digital image forensics. This research presents a wavelet-based detection method that uses discrete wavelet transform (DWT) preprocessing and a ResNet50 classification layer to differentiate the StyleGAN-generated images from real ones. Haar and Daubechies wavelet filters are applied to convert the input images into multi-resolution representations, which will then be fed to a ResNet50 network for classification, capitalizing on subtle artifacts left by the generative process. Moreover, the wavelet-based models are compared to an identical ResNet50 model trained on spatial data. The Haar and Daubechies preprocessed models achieved a greater accuracy of 93.8 percent and 95.1 percent, much higher than the model developed in the spatial domain (accuracy rate of 81.5 percent). The Daubechies-based model outperforms Haar, showing that adding layers of descriptive frequency patterns can lead to even greater distinguishing power. These results indicate that the GAN-generated images have unique wavelet-domain artifacts or "fingerprints." The method proposed illustrates the effectiveness of wavelet-domain analysis to detect GAN images and emphasizes the potential of further developing the capabilities of future deepfake detection systems.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。