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 2106.14191cs.CRcs.SYeess.SY
收藏

芝麻开门将访问控制引入语音服务

Open, Sesame! Introducing Access Control to Voice Services

Dominika Woszczyk, Alvin Lee, Soteris Demetriou

摘要:

个人语音助理(VA)被证明容易受到记录和重放以及其他声学攻击的攻击,这些攻击允许对手获得对灵巧家庭内连接设备的未经授权的控制。现有的防御要么缺乏检测和管理功能,要么过于粗粒度,无法实现与其他计算接口同等的灵活策略。在这项工作中,我们提出了Sesame,这是一个用于边缘设备的轻量级框架,它是第一个实现智能家居语音命令细粒度访问控制的框架。Sesame结合了三个组件:自动语音识别、自然语言理解(NLU)和策略模块。我们在Android设备上实现了Sesame,并证明了我们的系统可以实时(362ms端到端推理时间)执行Alexa和Google Home的安全策略,使用轻量级(<25MB)NLU模型,与非紧凑型模型相比,其准确性损失最小。

英文摘要:

Personal voice assistants (VAs) are shown to be vulnerable against record-and-replay, and other acoustic attacks which allow an adversary to gain unauthorized control of connected devices within a smart home. Existing defenses either lack detection and management capabilities or are too coarse-grained to enable flexible policies on par with other computing interfaces. In this work, we present Sesame, a lightweight framework for edge devices which is the first to enable fine-grained access control of smart-home voice commands. Sesame combines three components: Automatic Speech Recognition, Natural Language Understanding (NLU) and a Policy module. We implemented Sesame on Android devices and demonstrate that our system can enforce security policies for both Alexa and Google Home in real-time (362ms end-to-end inference time), with a lightweight (<25MB) NLU model which exhibits minimal accuracy loss compared to its non-compact equivalent.

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。

↑