通过中间层嵌入敏感性检测AI生成图像，效果优于现有方法。
Intermediate Representations are Strong AI-Generated Image Detectors
- 利用图像在中间层的嵌入对扰动的敏感性进行检测。
- 在Forensics Small数据集上比最优无训练方法提升39.61% AUROC。
- 无需训练即可跨域适用，适合快速部署的图像真伪检测场景。
生成式AI模型的快速发展使得逼真图像的生成成为可能，但随之而来的是对内容滥用的担忧和对有效检测技术的迫切需求。当前基于训练的检测方法通常计算成本高，难以泛化到未见数据域；而无训练方法则检测性能不足。为此，我们提出一种基于搜索的方法，利用中间层数据嵌入对扰动的敏感性来检测AI生成图像。给定真实与生成图像集合，该方法通过比较原始图像嵌入与扰动后嵌入的相似性来判断图像来源。我们在两个综合性基准测试GenImage和Forensics Small上验证了该方法。相比现有的无训练与有训练的最先进方法，本方法在不同数据集上均表现出更优性能。平均而言，在Forensics Small基准上，其AUROC得分较最优无训练方法提升39.61%，较最优有训练方法提升5.14%。
原文摘要 · Abstract (English)
The rapid advancement in generative AI models has enabled the creation of photorealistic images. At the same time, there are growing concerns about the potential misuse and dangers of generated content, as well as a pressing need for effective AI-generated image detectors. However, current training-based detection techniques are typically computationally costly and can hardly be generalized to unseen data domains, while training-free methods fall short in detection performance. To bridge this gap, we propose a search-based method employing data embedding sensitivity in intermediate layers to detect AI-generated images. Given a set of real and AI-generated images, our method examines the similarity between original image embeddings and perturbed image embeddings, and detects AI-generated images based on the similarity. We examine the proposed method on two comprehensive benchmarks: GenImage and Forensics Small. Our method exhibits improved performance across different datasets compared to both training-free and training-based state-of-the-art methods. On average, our method achieves the largest performance gain on the Forensics Small benchmark by 39.61% compared to the best training-free method and 5.14% compared to the best training-based method in AUROC score.
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。