注意力捕获并非检测：人类如何错过AI图像局部编辑的两阶段解释
Attention Capture Is Not Detection: A Two-Stage Account of How Humans Miss Localized AI Image Edits
浏览论文内容
中文总结 AI 辅助
该研究通过眼动实验发现人类对AI图像编辑的感知分为注意力捕获和判断两个独立阶段，训练生成扫描路径的Transformer可有效建模注意力捕获阶段，为遏制AI虚假信息提供新视角。
中文摘要 AI 辅助
随着AI生成的图像编辑内容激增，旨在遏制由此产生的虚假信息的平台将可检测性视为单一、无差别的属性：编辑要么得到警告，要么没有。我们表明这是错误的模型。在一项受控的眼动追踪研究中（N=59，拉丁方设计，四个条件交叉编辑区域与语义合理性），混合效应分析显示，编辑是否被注意到以及是否被正确判断为虚假是可分离的阶段，由不同因素主导：编辑区域驱动注意力捕获（p<0.001），而语义合理性驱动判断准确性和“看到但未看见（LBFS）”错误率（p<0.001）。这种分离在多重比较校正后仍然存在；两个因素之间的二次交互作用则不存在。这一两阶段解释将视觉注意力研究中长期存在的区分（前注意捕获与费力识别之间的区分）扩展到AI编辑可检测性的新领域。我们随后测试了生成式眼动模型是否可在计算上实现注意力捕获阶段：经过训练生成扫描路径的Transformer以强判别力追踪每幅图像的注意力（在保留的测试刺激中皮尔逊r=0.77至0.82），在预测LBFS发生率这一更困难的任务上，即使不访问合理性标签，也适度优于两参数线性基线（r=0.52对比r=0.48）。我们如实报告了这一比较、消融实验及方法的局限性（单一固定的训练/验证拆分，非留一被试交叉验证），以负责任地传达机器学习系统在遏制AI驱动虚假信息方面的能力与不足。
英文摘要
As AI-generated image edits proliferate, the platforms meant to curb the resulting disinformation treat detectability as a single, undifferentiated property: an edit either gets a warning or it does not. We show this is the wrong model. Across a controlled eye-tracking study ($N=59$, Latin-square design, four conditions crossing edit area and semantic plausibility), a mixed-effects analysis reveals that whether an edit is noticed and whether it is correctly judged as fake are dissociable stages, governed by different factors: edit area drives attention capture ($p<0.001$) while semantic plausibility drives judgment accuracy and look-but-fail-to-see (LBFS) error rates ($p<0.001$). This dissociation survives correction for multiple comparisons; a secondary interaction between the two factors does not. This two-stage account extends a long-standing distinction in visual attention research (between pre-attentive capture and effortful recognition) into the new domain of AI-edit detectability. We then test whether a generative eye-movement model can computationally operationalize the attention-capture stage: a Transformer trained to generate scanpaths tracks per-image attention with strong discriminative power (Pearson $r=0.77$--$0.82$ across held-out stimuli) and, on the harder task of predicting LBFS incidence, modestly outperforms a two-parameter linear baseline even without access to the plausibility label ($r=0.52$ vs. $r=0.48$). We report this comparison, our ablations, and our method's limitations (a single fixed train/validation split, not leave-one-subject-out) without inflation, consistent with responsibly communicating what a machine learning system can and cannot do to help curb AI-driven disinformation.
发表机构
- National Chengchi University（国立政治大学）
机构由 AI 辅助整理，请以论文原文为准。
