DiffusionPrint: 为基于扩散的修复定位学习生成指纹
DiffusionPrint: Learning Generative Fingerprints for Diffusion-Based Inpainting Localization
AI总结：
本文提出DiffusionPrint，通过对比学习框架学习鲁棒于潜在解码引入的频谱畸变的取证信号，提升基于融合的图像伪造定位性能，实验表明在多个生成模型上均取得显著提升。
摘要：
现代基于扩散的修复模型对图像伪造定位（IFL）提出了重大挑战，因为其完整的再生管道通过潜在解码器重建整个图像，破坏了现有取证方法依赖的相机级噪声模式。我们提出了DiffusionPrint，一种基于补丁的对比学习框架，学习一种鲁棒于潜在解码引入的频谱畸变的取证信号。它利用了由相同模型生成的修复区域具有一致的生成指纹这一事实，将其用作自监督信号。DiffusionPrint通过MoCo式目标训练卷积主干，结合跨类别难负样本挖掘和生成器感知的分类头，产生一个用作融合基于IFL框架中高度判别性二次模态的取证特征图。集成到TruFor、MMFusion和轻量级融合基线中，DiffusionPrint在多个生成模型上一致提升定位性能，在未见的掩码类型上获得高达+28%的提升，并已验证在未见的生成架构上具有通用性。代码可在https://github.com/mever-team/diffusionprint获取。
英文摘要：
Modern diffusion-based inpainting models pose significant challenges for image forgery localization (IFL), as their full regeneration pipelines reconstruct the entire image via a latent decoder, disrupting the camera-level noise patterns that existing forensic methods rely on. We propose DiffusionPrint, a patch-level contrastive learning framework that learns a forensic signal robust to the spectral distortions introduced by latent decoding. It exploits the fact that inpainted regions generated by the same model share a consistent generative fingerprint, using this as a self-supervisory signal. DiffusionPrint trains a convolutional backbone via a MoCo-style objective with cross-category hard negative mining and a generator-aware classification head, producing a forensic feature map that serves as a highly discriminative secondary modality in fusion-based IFL frameworks. Integrated into TruFor, MMFusion, and a lightweight fusion baseline, DiffusionPrint consistently improves localization across multiple generative models, with gains of up to +28% on mask types unseen during fine-tuning and confirmed generalization to unseen generative architectures. Code is available at https://github.com/mever-team/diffusionprint
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
