arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2603.28092cs.LG
收藏

InkDrop: 针对数据集压缩的不可见后门攻击

InkDrop: Invisible Backdoor Attacks Against Dataset Condensation

IEEE Publication Technology Group(IEEE出版技术组)

He Yang, Dongyi Lv, Song Ma, Wei Xi, Zhi Wang, Hanlin Gu, Yajie Wang

AI总结:

本文提出InkDrop,一种针对数据集压缩的隐蔽后门攻击方法,通过利用模型决策边界附近的不确定性,实现对压缩数据集的高效且隐蔽的恶意操控。

摘要:

数据集压缩(DC)是一种数据高效的学习范式,通过合成小型但信息丰富的数据集,使模型能够达到全数据训练的性能。然而,近期的研究揭示了DC对后门攻击的关键漏洞,恶意模式(例如触发器)被植入到压缩数据集中,导致特定输入的针对性误分类。现有攻击总是优先考虑攻击效果和模型实用性,忽略了隐蔽性这一关键维度。为弥合这一差距,我们提出了InkDrop,通过增强恶意操控的不可见性,同时不降低攻击效果和模型实用性。InkDrop利用模型决策边界附近固有的不确定性,其中微小的输入扰动可引起语义变化,以构建隐蔽且有效的后门攻击。具体而言,InkDrop首先选择接近目标决策边界的候选样本,这些样本具有潜在的语义亲和力与目标类别。然后,它学习实例依赖的扰动,受感知和空间一致性约束,将目标恶意行为嵌入到压缩数据集中。在多种数据集上的广泛实验验证了InkDrop的整体有效性,证明其能够将对抗意图整合到压缩数据集中,同时保持模型实用性并最小化可检测性。我们的代码可在https://github.com/lvdongyi/InkDrop上获得。

英文摘要:

Dataset Condensation (DC) is a data-efficient learning paradigm that synthesizes small yet informative datasets, enabling models to match the performance of full-data training. However, recent work exposes a critical vulnerability of DC to backdoor attacks, where malicious patterns (\textit{e.g.}, triggers) are implanted into the condensation dataset, inducing targeted misclassification on specific inputs. Existing attacks always prioritize attack effectiveness and model utility, overlooking the crucial dimension of stealthiness. To bridge this gap, we propose InkDrop, which enhances the imperceptibility of malicious manipulation without degrading attack effectiveness and model utility. InkDrop leverages the inherent uncertainty near model decision boundaries, where minor input perturbations can induce semantic shifts, to construct a stealthy and effective backdoor attack. Specifically, InkDrop first selects candidate samples near the target decision boundary that exhibit latent semantic affinity to the target class. It then learns instance-dependent perturbations constrained by perceptual and spatial consistency, embedding targeted malicious behavior into the condensed dataset. Extensive experiments across diverse datasets validate the overall effectiveness of InkDrop, demonstrating its ability to integrate adversarial intent into condensed datasets while preserving model utility and minimizing detectability. Our code is available at https://github.com/lvdongyi/InkDrop.

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。

↑