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 2410.01697cs.LGcs.CV
收藏

通过多目标表征学习增强对抗鲁棒性

Enhancing Adversarial Robustness through Multi-Objective Representation Learning

Safe Autonomous Systems, TU Dortmund University, Germany(安全自主系统,图恩多特大学,德国);Lamarr Institute for Machine Learning and Artificial Intelligence(拉马尔机器学习与人工智能研究所)

Sedjro Salomon Hotegni, Sebastian Peitz

AI总结:

针对DNNs易受对抗扰动的问题,提出多目标方法MOREL,利用余弦相似度和多正例对比损失对齐自然与对抗特征,显著提升了模型对白盒和黑盒攻击的鲁棒性。

摘要:

深度神经网络(DNNs)容易受到微小对抗扰动的影响,这些对输入数据的微小改变看似微不足道,却会导致模型产生截然不同的输出。许多防御方法需要在评估期间修改模型架构或执行测试时数据净化。这不仅引入了额外的复杂性,而且通常依赖于特定架构。然而我们表明,在训练期间进行鲁棒特征学习可以显著增强DNN的鲁棒性。我们提出了MOREL,这是一种多目标方法,利用余弦相似度和多正例对比损失对齐自然特征和对抗特征,以鼓励同类输入产生相似特征。大量实验表明,MOREL显著提升了对白盒和黑盒攻击的鲁棒性。代码可在 https://github.com/salomonhotegni/MOREL 获取。

英文摘要:

Deep neural networks (DNNs) are vulnerable to small adversarial perturbations, which are tiny changes to the input data that appear insignificant but cause the model to produce drastically different outputs. Many defense methods require modifying model architectures during evaluation or performing test-time data purification. This not only introduces additional complexity but is often architecture-dependent. We show, however, that robust feature learning during training can significantly enhance DNN robustness. We propose MOREL, a multi-objective approach that aligns natural and adversarial features using cosine similarity and multi-positive contrastive losses to encourage similar features for same-class inputs. Extensive experiments demonstrate that MOREL significantly improves robustness against both white-box and black-box attacks. Our code is available at https://github.com/salomonhotegni/MOREL

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。

↑