Acoustic Sensing: Mobile Applications and Frameworks

[摘要] Acoustic sensing has attracted significant attention from both academia and industry due to its ubiquity. Since smartphones and many IoT devices are already equipped with microphones and speakers, it requires nearly zero additional deployment cost. Acoustic sensing is also versatile. For example, it can detect obstacles for distracted pedestrians (BumpAlert), remember indoor locations through recorded echoes (EchoTag), and also understand the touch force applied to mobile devices (ForcePhone).In this dissertation, we first propose three acoustic sensing applications, BumpAlert, EchoTag, and ForcePhone, and then introduce a cross-platform sensing framework called LibAS. LibAS is designed to facilitate the development of acoustic sensing applications. For example, LibAS can let developers prototype and validate their sensing ideas and apps on commercial devices without the detailed knowledge of platform-dependent programming. LibAS is shown to require less than 30 lines of code in Matlab to implement the prototype of ForcePhone on Android/iOS/Tizen/Linux devices.

[发布机构] University of Michigan

[效力级别] mobile computing

[关键词] acoustic sensing;mobile computing;Computer Science;Engineering (General);Engineering;Science;Computer Science & Engineering