arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2603.12915cs.CVcs.AI
收藏

标注点:结构忠实的实例删除

Stake the Points: Structure-Faithful Instance Unlearning

Chung-Ang University(Chung-Ang 大学)

Kiseong Hong, JungKyoo Shin, Eunwoo Kim

AI总结:

本文提出结构忠实的实例删除框架,通过引入语义锚点保持知识结构,提升图像分类、检索和人脸识别性能,实现删除与保留的平衡。

摘要:

机器删除(MU)旨在解决预训练模型的隐私风险。其目标是删除指定数据的影响,同时保持保留知识的效用。实现这一目标需要保持保留实例间的语义关系,而现有研究常忽视这一点。我们观察到,没有这种保持,模型会经历逐步的结构崩溃,破坏删除与保留的平衡。在本文中,我们提出了一种新的结构忠实框架,引入了“标注点”,即作为参考点的语义锚点,以维持知识结构。通过利用这些锚点,我们的框架捕捉并稳定了知识的语义组织。具体来说,我们从语言驱动的属性描述中实例化锚点,这些描述由语义编码器(如CLIP)编码。我们通过结构感知对齐和正则化来强制保持知识结构:前者在删除前后对保留知识的组织进行对齐,后者调节对结构关键参数的更新。在图像分类、检索和人脸识别中的结果表明,性能平均提升了32.9%、22.5%和19.3%,在平衡删除与保留的权衡的同时增强了泛化能力。

英文摘要:

Machine unlearning (MU) addresses privacy risks in pretrained models. The main goal of MU is to remove the influence of designated data while preserving the utility of retained knowledge. Achieving this goal requires preserving semantic relations among retained instances, which existing studies often overlook. We observe that without such preservation, models suffer from progressive structural collapse, undermining both the deletion-retention balance. In this work, we propose a novel structure-faithful framework that introduces stakes, i.e., semantic anchors that serve as reference points to maintain the knowledge structure. By leveraging these anchors, our framework captures and stabilizes the semantic organization of knowledge. Specifically, we instantiate the anchors from language-driven attribute descriptions encoded by a semantic encoder (e.g., CLIP). We enforce preservation of the knowledge structure via structure-aware alignment and regularization: the former aligns the organization of retained knowledge before and after unlearning around anchors, while the latter regulates updates to structure-critical parameters. Results from image classification, retrieval, and face recognition show average gains of 32.9%, 22.5%, and 19.3% in performance, balancing the deletion-retention trade-off and enhancing generalization.

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。

↑