探索类增量持续学习中的跨阶段对抗转移性
Exploring Cross-Stage Adversarial Transferability in Class-Incremental Continual Learning
AI总结：
本文探讨了类增量持续学习中对抗转移攻击的脆弱性，揭示了持续学习方法对这类攻击的高度敏感性，并指出现有防御方法的不足。
摘要：
类增量持续学习通过使分类模型在获取新类别的同时保留之前学习类别的知识来解决灾难性遗忘问题。然而，这一过程中模型对对抗攻击的脆弱性尚未得到充分研究。在本文中，我们首次探讨了对阶段转移攻击的脆弱性，即使用早期阶段的模型生成的对抗示例用于攻击后期阶段的模型。我们的发现表明，持续学习方法对这些攻击极为敏感，提出了严重的安全问题。我们通过阶段间的模型相似性和逐渐的鲁棒性退化现象解释了这一现象。此外，我们发现现有的基于对抗训练的防御方法对阶段转移攻击并不足够有效。代码可在 https://github.com/mcml-official/CSAT 上获取。
英文摘要：
Class-incremental continual learning addresses catastrophic forgetting by enabling classification models to preserve knowledge of previously learned classes while acquiring new ones. However, the vulnerability of the models against adversarial attacks during this process has not been investigated sufficiently. In this paper, we present the first exploration of vulnerability to stage-transferred attacks, i.e., an adversarial example generated using the model in an earlier stage is used to attack the model in a later stage. Our findings reveal that continual learning methods are highly susceptible to these attacks, raising a serious security issue. We explain this phenomenon through model similarity between stages and gradual robustness degradation. Additionally, we find that existing adversarial training-based defense methods are not sufficiently effective to stage-transferred attacks. Codes are available at https://github.com/mcml-official/CSAT.
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。
