Abstract
Recent advances in generative image modeling have achieved visual realism sufficient to deceive human experts, yet their potential for privacy preserving data sharing remains insufficiently understood. A central obstacle is the absence of reliable memorization detection mechanisms, limited quantitative evaluation, and poor generalization of existing privacy auditing methods across domains. To address this, we propose to view memorization detection as a unified problem at the intersection of re-identification and copy detection, whose complementary goals cover both identity consistency and augmentation-robust duplication, and introduce Latent Contrastive Memorization Network (LCMem), a cross-domain model evaluated jointly on both tasks. LCMem achieves this through a two-stage training strategy that first learns identity consistency before incorporating augmentation-robust copy detection. Across six benchmark datasets, LCMem achieves improvements of up to 16 percentage points on re-identification and 30 percentage points on copy detection, enabling substantially more reliable memorization detection at scale. Our results show that existing privacy filters provide limited performance and robustness, highlighting the need for stronger protection mechanisms. We show that LCMem sets a new standard for cross-domain privacy auditing, offering reliable and scalable memorization detection. Code and model is publicly available at this https URL.
Abstract (translated)
近期在生成图像模型方面的进展已经达到了足以欺骗人类专家的视觉逼真度，然而这些技术在保护隐私的数据共享中的潜力仍然不够充分地被理解。一个主要障碍是缺乏可靠的内存检测机制、有限的定量评估以及现有隐私审计方法跨领域的差泛化能力。为了解决这个问题，我们建议将内存检测视为识别和复制检测交叉点上的统一问题，其互补目标涵盖了身份一致性与增强鲁棒性复制，并引入了潜对比记忆网络（LCMem），这是一种在两个任务上联合评估的跨领域模型。 LCMem 通过两阶段训练策略实现这一目标，首先学习身份一致性，然后将增强鲁棒性复制检测纳入其中。在六个基准数据集上，LCMem 在重新识别方面提高了多达16个百分点，在复制检测方面提高了30个百分点，从而实现了更大规模下更可靠的内存检测。我们的结果显示现有的隐私过滤器提供的性能和稳健性有限，强调了需要更强的保护机制。我们证明 LCMem 为跨域隐私审计设定了新的标准，提供了可靠且可扩展的记忆检测方法。代码和模型在以下链接公开提供：[此 URL]（请将括号内的文本替换为实际发布的URL）。
URL
https://arxiv.org/abs/2512.14421
