无法窃取我的人脸：针对身份保留个性化生成的主动防御
No Way To Steal My Face: Proactive Defense Against Identity-Preserving Personalized Generation
AI总结：
针对扩散模型个性化生成的人脸隐私问题，提出可防御两类个性化方法的IDGuardian框架，通过破坏身份提取与注入阶段实现身份保护，且具备鲁棒性与泛化性。
摘要：
扩散模型的最新进展已能为数字化身、虚拟试穿系统等个性化应用生成高保真且保留身份的图像。然而，这些模型依赖敏感的人脸参考图像，引发了日益增长的隐私担忧。现有防御机制主要针对基于训练的个性化设计，由于其身份整合范式存在根本差异，难以推广到新兴的无训练方法。为弥合这一差距，我们提出IDGuardian——首个可泛化且与模型无关的身份保护框架，能够防御基于训练和无训练的两种个性化方法。IDGuardian将个性化过程抽象为两个关键阶段：身份提取与身份注入，随后引入精心设计的对抗扰动以同时破坏这两个阶段。具体而言，它会降低外部编码器提取的身份特征质量，并建立一条对抗性概念桥梁，使生成轨迹偏离目标身份。大量实验表明，IDGuardian可在各种个性化流程和模型架构中有效保护身份，同时对后处理、自适应攻击及跨数据集泛化保持鲁棒性。我们的代码可在https://github.com/OpenAscent-L/IDGuardian.git获取。
英文摘要：
Recent advances in diffusion models have enabled high-fidelity, identity-preserving image generation for personalized applications such as digital avatars and virtual try-on systems. However, their reliance on sensitive facial reference images raises growing privacy concerns. Existing defense mechanisms are primarily designed for training-based personalization and struggle to generalize to emerging training-free approaches, due to fundamental differences in their identity integration paradigms. To bridge this gap, we propose IDGuardian-the first generalizable and model-agnostic identity protection framework capable of defending against both training-based and training-free personalization methods. IDGuardian abstracts the personalization process into two critical stages: identity extraction and identity injection. It then introduces crafted adversarial perturbations to simultaneously disrupt both stages. Specifically, it degrades the identity features extracted by external encoders and establishes an adversarial conceptual bridge that misdirects the generative trajectory away from the target identity. Extensive experiments show that IDGuardian effectively protects identity across various personalization pipelines and model architectures, while remaining robust to post-processing, adaptive attacks, and cross-dataset generalization. Our code is available at https://github.com/OpenAscent-L/IDGuardian.git.
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。
