Biologically Inspired Acoustic Source Localization System

This proposal is to develop the acoustic source detection and localization algorithms identified as optimal in Phase I into a wearable prototype that integrates a networked sensor array, digital audio processing, position and orientation tracking of the wearer, and aural/visual displays of the detected sounds. Advantages of our approach over competing algorithms include noise tolerance, power efficiency, multi-source detection and localization, better-than-human localization accuracy, and effective display techniques. The system will be designed to be compatible with and easily integrated into other wearable audio technologies such as transparent hearing, noise protection, and supernormal hearing systems. The system will use detection and localization algorithms developed by leading acoustics researchers at Virginia Polytechnic Institute and State University and the University of Wisconsin, which proved very effective in Phase I. The algorithms will be implemented in software by AuSIM’s engineering team. The sensors, digital processing, and display devices will be integrated into a miniaturized, wearable system. The prototypes will then be linked into a network and tested in various environments representative of expected end-user situations.