Defending against Thru-barrier Stealthy Voice Attacks via Cross-Domain Sensing on Phoneme Sounds

Defending against Thru-barrier Stealthy Voice Attacks via Cross-Domain Sensing on Phoneme Sounds
复制标题

通过音素声音的跨域感知防御穿墙隐形语音攻击

DOI:
10.1109/icdcs54860.2022.00071
复制
发表时间:
2022
期刊:
2022 IEEE 42nd International Conference on Distributed Computing Systems (ICDCS
影响因子:
--
通讯作者:
Chen, Yingying
Chen, Yingying
中科院分区:
--
文献类型:
--
作者:
Shi, Cong;Zhao, Tianming;Zhang, Wenjin;Mahdad, Ahmed Tanvir;Ye, Zhengkun;Wang, Yan;Saxena, Nitesh;Chen, Yingying

文献摘要

参考文献

相似文献

语音输入的开放性质使得语音助理(VA)系统易受各种声学攻击(例如,重放和语音合成攻击)。对手发起这些攻击的一种简单而有效的方法是躲在障碍物后面(例如,墙、窗或门),并发出未经授权的语音命令而不被合法用户观察到。在这项工作中,我们开发了一种自动化、无需训练的防御系统,可以保护VA系统免受此类透屏障声学攻击。我们的研究发现,声信号通过障碍物一般呈现出独特的频率选择性的影响,在振动域。因此,我们建议设计一个系统,通过利用用户可穿戴设备中的低成本、跨域传感来捕获这种独特的障碍效应。该系统通过可穿戴设备的扬声器重放音频域信号,并通过内置加速度计捕获由振动域中的音频声音引起的传导振动。为了提高该系统的可靠性,我们开发了一种独特的振动域增强方法来提取最敏感的障碍物的频率选择性效果的音素。我们确定有效的振动域功能,捕捉障碍的影响,在振动域。开发了一种基于2D相关性的方法来检查VA系统和用户可穿戴设备的录音之间的语音相似性,并检测穿透屏障攻击。在各种障碍和环境下进行的大量实验表明,所提出的防御系统可以有效地防御随机,重放,合成和隐藏的语音攻击,平均错误率低于4%。
The open nature of voice input makes voice assistant (VA) systems vulnerable to various acoustic attacks (e.g., replay and voice synthesis attacks). A simple yet effective way for adversaries to launch these attacks is to hide behind barriers (e.g., a wall, a window, or a door) and give unauthorized voice commands without being observed by legitimate users. In this work, we develop an automated, training-free defense system that can protect VA systems from such thru-barrier acoustic attacks. Our study finds that acoustic signals passing through the barriers generally present a unique frequency-selective effect in the vibration domain. Thus, we propose to devise a system to capture this unique effect of barriers by leveraging low-cost, cross-domain sensing available in users’ wearables. The system replays the audio-domain signals with the wearable’s speaker and captures the conductive vibrations caused by the audio sounds in the vibration domain via the built-in accelerometer. To improve the proposed system’s reliability, we develop a unique vibration-domain enhancement method to extract the phonemes most sensitive to the frequency-selective effect of barriers. We identify effective vibration-domain features that capture the barriers’ effects in the vibration domain. A 2D-correlation-based method is developed to examine the speech similarity between the recordings from the VA system and the user’s wearable and detect thru-barrier attacks. Extensive experiments with various barriers and environments demonstrate that the proposed defense system can effectively defend random, replay, synthesis, and hidden voice attacks with less than 4% equal error rates.
DOI: 10.1109/icccn49398.2020.9209697
发表时间: 2020-08
期刊: 2020 29th International Conference on Computer Communications and Networks (ICCCN)
影响因子: --
作者:
Cong Shi;Xiaonan Guo;Ting Yu;Yingying Chen;Yucheng Xie;Jian Liu
通讯作者: Cong Shi;Xiaonan Guo;Ting Yu;Yingying Chen;Yucheng Xie;Jian Liu
使用自适应频谱子带质心进行扬声器验证
DOI: --
发表时间: 2007
期刊: International Conference on Biometrics
影响因子: --
作者:
T. Kinnunen;Bingjun Zhang;Jia Zhu;Ye Wang
通讯作者: Ye Wang
DOI: --
发表时间: 2016
期刊: Asia Pacific Conference on Circuits and Systems
影响因子: --
作者:
Po;C. Yeh;H. Tsai;Y. Juang
通讯作者: Y. Juang
DOI: 10.1145/3427228.3427259
发表时间: 2020-12
期刊: Proceedings of the 36th Annual Computer Security Applications Conference
影响因子: --
作者:
Cong Shi;Yan Wang;Yingying Chen;Nitesh Saxena;Chen Wang
通讯作者: Cong Shi;Yan Wang;Yingying Chen;Nitesh Saxena;Chen Wang
(三)《日本研究与词语遗漏》反思16世纪吴语语音学
DOI: --
发表时间: 2004
期刊: 海外事情研究(熊本学園大学海外事情研究所) 第32巻第1巻
影响因子: --
作者:
丁 鋒;丁 鋒;丁 鋒;丁 鋒;丁 鋒;丁鋒;丁鋒
通讯作者: 丁鋒