融合语义与像素特征，提升AI生成图像检测的泛化能力
CO-SPY: Combining Semantic and Pixel Features to Detect Synthetic Images by AI
- 结合手部指节数等语义特征与像素差异等伪影特征
- 在22个生成模型上平均准确率提升11%至34%
- 适合关注生成图像安全检测的研究者与应用开发者
随着生成式AI的快速发展，如今可在数秒内合成高质量图像。尽管技术强大，但其滥用引发严重担忧。现有真实与生成图像区分方法泛化能力不足，仅对特定生成模型有效，且易受JPEG压缩等后处理技术干扰。为此，我们提出新框架Co-Spy，先增强现有语义特征（如手部指节数）和伪影特征（如像素值差异），再自适应融合，实现更通用、更鲁棒的合成图像检测。此外，我们构建了Co-Spy-Bench数据集，包含5个真实图像数据集和22个前沿生成模型（含最新FLUX模型），并从互联网收集5万张真实世界中的合成图像，以支持更贴近实际的评估。大量实验表明，在相同训练条件下，本检测器性能优于现有方法，平均准确率提升约11%至34%。代码已公开于https://github.com/Megum1/Co-Spy。
原文摘要 · Abstract (English)
With the rapid advancement of generative AI, it is now possible to synthesize high-quality images in a few seconds. Despite the power of these technologies, they raise significant concerns regarding misuse. Current efforts to distinguish between real and AI-generated images may lack generalization, being effective for only certain types of generative models and susceptible to post-processing techniques like JPEG compression. To overcome these limitations, we propose a novel framework, Co-Spy, that first enhances existing semantic features (e.g., the number of fingers in a hand) and artifact features (e.g., pixel value differences), and then adaptively integrates them to achieve more general and robust synthetic image detection. Additionally, we create Co-Spy-Bench, a comprehensive dataset comprising 5 real image datasets and 22 state-of-the-art generative models, including the latest models like FLUX. We also collect 50k synthetic images in the wild from the Internet to enable evaluation in a more practical setting. Our extensive evaluations demonstrate that our detector outperforms existing methods under identical training conditions, achieving an average accuracy improvement of approximately 11% to 34%. The code is available at https://github.com/Megum1/Co-Spy.
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。