arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2411.09220eess.AScs.AIeess.SP
收藏

针对ASR的可迁移对抗攻击

Transferable Adversarial Attacks against ASR

Institute for Infocomm Research, Agency for Science, Technology, and Research (A*STAR)(新加坡科技研究局信息通信研究院);National University of Singapore(新加坡国立大学);University of California Riverside(加州大学河滨分校);Shenzhen Research Institute of Big Data, The Chinese University of Hong Kong(香港中文大学(深圳)深圳大数据研究院)

Xiaoxue Gao, Zexin Li, Yiming Chen, Cong Liu, Haizhou Li

AI总结:

本文研究黑盒ASR模型的鲁棒性,提出结合可微特征提取器的两种时域可迁移攻击,并设计语音感知梯度优化方法(SAGO)以在保持人类不可感知性的同时迫使错误转录,在多模型与数据库上验证了有效性。

摘要:

鉴于自动语音识别(ASR)的广泛研究与实际应用,确保ASR模型对微小输入扰动的鲁棒性成为维持其在实时场景中有效性的关键考量。以往对ASR模型鲁棒性的探索主要集中在拥有ASR模型完全访问权限的白盒设置下评估准确性。然而,在实际应用中,完整的ASR模型细节通常不可获取。因此,评估黑盒ASR模型的鲁棒性对于全面理解ASR模型的韧性至关重要。对此,我们深入研究了前沿ASR模型在实际黑盒攻击下的脆弱性,并提出采用两种先进的基于时域的可迁移攻击,结合我们提出的可微特征提取器。我们还为ASR提出了一种语音感知梯度优化方法(SAGO),该方法通过语音活动检测规则和语音感知梯度导向优化器,在迫使错误转录的同时,将对人类不可感知性的影响降至最低。综合实验结果表明,与基线方法相比,在两个数据库的五个模型上均取得了性能提升。

英文摘要:

Given the extensive research and real-world applications of automatic speech recognition (ASR), ensuring the robustness of ASR models against minor input perturbations becomes a crucial consideration for maintaining their effectiveness in real-time scenarios. Previous explorations into ASR model robustness have predominantly revolved around evaluating accuracy on white-box settings with full access to ASR models. Nevertheless, full ASR model details are often not available in real-world applications. Therefore, evaluating the robustness of black-box ASR models is essential for a comprehensive understanding of ASR model resilience. In this regard, we thoroughly study the vulnerability of practical black-box attacks in cutting-edge ASR models and propose to employ two advanced time-domain-based transferable attacks alongside our differentiable feature extractor. We also propose a speech-aware gradient optimization approach (SAGO) for ASR, which forces mistranscription with minimal impact on human imperceptibility through voice activity detection rule and a speech-aware gradient-oriented optimizer. Our comprehensive experimental results reveal performance enhancements compared to baseline approaches across five models on two databases.

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。

↑