利用扩散模型注意力图谱的频谱特征，无训练检测图像复制粘贴伪造。
Spectral Forensics of Diffusion Attention Graphs for Copy-Move Forgery Detection
- 通过分析预训练扩散模型中注意力图的频谱结构发现伪造痕迹。
- 在多个数据集上达到最高0.774的AUROC，CoMoFoD上达0.833 AUPRC。
- 无需微调，适合图像完整性验证与数字取证研究者使用。
复制粘贴伪造（copy-move forgery）通过复制图像局部区域来篡改内容，严重威胁视觉媒体真实性。本文提出GraphSpecForge，一种无需训练的检测框架，通过分析预训练Stable Diffusion U-Net的自注意力图谱的谱结构来识别伪造。核心洞察是：复制粘贴操作会在注意力图中引入近似子图重复，导致归一化图拉普拉斯矩阵的谱分布发生可测量的变化。我们基于扰动理论建立该关联，并采用图像级异常检测器，通过计算每张图像拉普拉斯谱与真实分布之间的Wasserstein距离进行判定。在四个基准数据集上评估，未进行伪造相关微调。在RecodAI-LUC（5,128张图像）上，最优配置达到AUROC = 0.606（95%置信区间：0.580–0.638；置换检验p=0.005），归一化拉普拉斯谱比原始注意力谱提升+0.057 AUROC。在MICC-F220、CoMoFoD和COVERAGE上分别获得AUROC 0.752、0.774、0.673；其中CoMoFoD还达到AUPRC = 0.833，平衡准确率0.712，MCC = 0.499，TPR@1%FPR = 32.5%。消融与伪造实验验证信号特异性及对篡改强度的敏感性，空图对照排除了平凡统计解释。
原文摘要 · Abstract (English)
Copy-move forgery, where a region within an image is duplicated to hide or fabricate content, remains a persistent threat to visual media integrity. We introduce GraphSpecForge, a training-free framework that detects copy-move forgery by analysing the spectral structure of attention graphs from a pretrained Stable Diffusion U-Net. Our central insight is that copy-move manipulation induces approximate subgraph duplication in the self-attention graph, leading to measurable spectral redistribution in the normalized graph Laplacian. We formalise this link with perturbation-based arguments and build an image-level anomaly detector using Wasserstein distances between per-image Laplacian spectra and an authentic reference distribution. We evaluate GraphSpecForge on four copy-move benchmarks without forgery-specific retraining. On RecodAI-LUC (5,128 images), our best configuration achieves AUROC = 0.606 (95% CI: 0.580-0.638; permutation p = 0.005), and the normalized Laplacian outperforms raw attention spectra by +0.057 AUROC. On MICC-F220, CoMoFoD, and COVERAGE, the same pipeline attains AUROCs of 0.752, 0.774, and 0.673, respectively; on CoMoFoD it also reaches AUPRC = 0.833, balanced accuracy = 0.712, MCC = 0.499, and TPR@1%FPR = 32.5%. Additional ablation and falsification experiments confirm the signal's specificity and sensitivity to manipulation strength, while null-graph controls rule out trivial-statistic explanations.
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。