层间一致性至关重要：用于可泛化合成图像检测的优雅潜变量转移差异
Layer Consistency Matters: Elegant Latent Transition Discrepancy for Generalizable Synthetic Image Detection
把预训练网络各层潜变量之间的特征转移一致性当作判别信号，可用于任何需要区分真实与生成数据的场景；这种跨层一致性度量也可迁移到深伪视频检测、图像取证与OOD检测中，替代依赖特定生成器伪影的低层统计线索。
原文依据 real images maintain consistent semantic attention and structural coherence in their latent representations, exhibiting more stable feature transitions across network layers
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论近期生成模型的快速进展显著提升了AI生成合成图像的逼真度与可获取性。在支撑各类创新应用的同时，这些合成图像前所未有的真实感使其与真实照片愈发难以区分，带来了媒体可信度与内容篡改等严重安全风险。尽管已有大量工作致力于合成图像检测，但多数现有方法依赖模型特定的伪影或低层统计线索，因而对未见数据的泛化能力较差。在本工作中，我们识别出一个此前未被探索的差异：真实图像在其潜变量表示中保持一致的语义注意力与结构连贯性，在跨网络层时表现出更稳定的特征转移，而合成图像则呈现出可辨识的独特模式。据此，我们提出一种新方法，称为潜变量转移差异（LTD），用于捕捉真实图像与合成图像在层间一致性上的差异。LTD 自适应地识别最具判别性的层，并评估跨层的转移差异。得益于所提出的层间判别建模，在包含多种 GAN 与扩散模型的三 个数据集上，我们的方法在平均准确率上超过基模型 14.35%。大量实验表明，LTD 优于近期最优方法，在检测准确率、泛化性与鲁棒性上均取得更优表现。
英文原文摘要 1541 字符
Recent rapid advancement of generative models has significantly improved the fidelity and accessibility of AI-generated synthetic images. While enabling various innovative applications, the unprecedented realism of these synthetics makes them increasingly indistinguishable from authentic photographs, posing serious security risks, such as media credibility and content manipulation. Although extensive efforts have been dedicated to detecting synthetic images, most existing approaches suffer from poor generalization to unseen data due to their reliance on model-specific artifacts or low-level statistical cues. In this work, we identify a previously unexplored distinction that real images maintain consistent semantic attention and structural coherence in their latent representations, exhibiting more stable feature transitions across network layers, whereas synthetic ones present discernible distinct patterns. Therefore, we propose a novel approach termed latent transition discrepancy (LTD), which captures the inter-layer consistency differences of real and synthetic images. LTD adaptively identifies the most discriminative layers and assesses the transition discrepancies across layers. Benefiting from the proposed inter-layer discriminative modeling, our approach exceeds the base model by 14.35% in mean Acc across three datasets containing diverse GANs and DMs. Extensive experiments demonstrate that LTD outperforms recent state-of-the-art methods, achieving superior detection accuracy, generalizability, and robustness.