发表机构
San Jose State University（圣何塞州立大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究针对生成器偏移问题，在SD3.5m数据集上评估了五个AI艺术检测器的鲁棒性，发现模型在跨生成器场景下泛化能力不足，CLIP ViT-L/14表现最优，为分层防御检测器的开发提供了依据。
AI 中文摘要
文本到图像生成模型发展迅速，现代扩散Transformer架构生成的图像越来越难以与人类创作的艺术品区分开。这一发展引发了对版权保护、虚假信息、欺诈、冒充以及数字内容真实性的重大担忧。大多数AI生成艺术检测器在同一生成器家族上进行训练和评估，而对新架构的鲁棒性尚未得到充分探索。在本研究中，我们基于稳定扩散3.5中等版本（SD3.5m）艺术品数据集分析生成器偏移，该数据集涵盖十种艺术风格，方法是对留出的人类艺术品样本进行反向提示。我们在基于U-Net的潜在扩散艺术品上训练了五个检测器，并在SD3.5m数据集的零样本跨生成器设置中进行评估。深度学习模型在分布内表现强劲，但在生成器偏移下会出现性能下降，将许多SD3.5m图像错误分类为人类创作，而人类创作图像的误报率仍然较低。CLIP ViT-L/14模型整体表现最佳，而Grad-CAM分析显示，假阴性的激活更弱且更分散。这些发现凸显了当前AI生成艺术检测器存在的泛化差距，推动将检测器开发为分层防御的一个组成部分，使其在快速发展的生成架构中保持可靠。
英文摘要
Text-to-image generative models have advanced rapidly, with modern Diffusion Transformer architectures producing images that are increasingly difficult to distinguish from human-created artwork. This development has raised significant concerns regarding copyright protection, misinformation, fraud, impersonation, and the authenticity of digital content. Most AI-art detectors are trained and evaluated on the same generator family, leaving robustness to newer architectures underexplored. In this chapter, we analyze generator shift based on a Stable Diffusion 3.5 Medium (SD3.5m) artwork dataset spanning ten art styles through reverse prompting of held-out human artwork samples. Five detectors are trained on U-Net-based latent diffusion artwork and evaluated in a zero-shot cross-generator setting on the SD3.5m dataset. Deep learning models perform strongly in-distribution but degrade under generator shift, misclassifying many SD3.5m images as human while human false positives remain low. The CLIP ViT-L/14 model performs best overall, while Grad-CAM analysis reveals weaker and more diffuse activation on false negatives. These findings highlight a generalization gap in current AI-art detectors and motivate the development of detectors as one component of a layered defense that remains reliable across rapidly evolving generative architectures.
CommentsTo appear as a chapter in the book "Artificial Intelligence for Cyber Defense in Emerging Threats", to be published by Springer by early 2027
