BioCase: Privacy Protection via Acoustic Sensing of Finger Touches on Smartphone Case Mini-Structures

Finger biometrics are widely used by smartphones as a secure and user-friendly credential for privacy protection. However, this information is difficult to measure without high-resolution images, leaving most works to treat this as an image-domain problem. We demonstrate that low-effort alternatives on smartphones are possible through the use of sound propagation in ubiquitous smartphone cases. Inexpensive and widely adopted, smartphone cases are always in contact with fingers, making them ideal for collecting finger biometrics. We thus design BioCase, an acoustic sensing system that leverages smartphone cases equipped with mini-structures to capture unique biometric-hybrid signatures (i.e., reflections influenced by the user's fingertip physiology and behavior) for smartphone privacy protection. The system generates inaudible structure-borne sound and measure the propagation through the smartphone case, mini-structures, and user finger. The design of the mini-structure controls the behavior of structure-borne sound such that unique responses are produced when different users and fingers touch the smartphone case. This enables low-cost, low-effort privacy protection, merely touching the smartphone case can authenticate users. Comprehensive experiments with 46 users over 10 weeks demonstrate BioCase can differentiate users with over 94% accuracy at a 5% false positive rate.

指纹生物识别技术被智能手机广泛用作一种安全且用户友好的隐私保护凭证。然而，如果没有高分辨率图像，这些信息很难测量，这使得大多数研究将其视为一个图像领域的问题。我们证明，通过利用智能手机普遍使用的手机壳中的声音传播，在智能手机上采用低成本的替代方法是可行的。智能手机壳价格低廉且被广泛使用，它们总是与手指接触，这使其成为收集指纹生物识别信息的理想选择。因此，我们设计了BioCase，这是一种声学传感系统，它利用配备微型结构的智能手机壳来捕获独特的生物识别混合特征（即受用户指尖生理和行为影响的反射），以实现智能手机的隐私保护。该系统产生听不见的结构传播的声音，并测量通过智能手机壳、微型结构和用户手指的传播情况。微型结构的设计控制了结构传播声音的行为，使得当不同用户和手指触摸智能手机壳时会产生独特的响应。这使得低成本、低投入的隐私保护成为可能，仅仅触摸智能手机壳就能对用户进行身份验证。在10周时间里对46名用户进行的综合实验表明，BioCase能够以5%的误报率达到超过94%的准确率来区分用户。