UROP Project

Using large Scale Microphone Arrays for Acoustic Source Localization and Geometric Reconstruction of Concert Halls

We want to estimate the geometry of concert halls using directional microphones. Large-Scale microphone arrays allow for a detailed analysis of room acoustics and the identification of individual sound waves as they move about the room. The data collected with such arrays makes it possible to track reflected sound as it bounces between the walls of concert halls. By combining the array elements in such a way that particular angles of sound incidence experience constructive interference (Beamforming techniques / phased array) a high-resolution directivity can be steered to search for individual reflections. Having found the time and the direction a reflected wave reaches the array allows estimating acoustic mirror image sources. Following the basic laws of geometric sound propagation it is possible to identify surfaces that reflect sound towards the array.

Task

* Understand how beamforming techniques work (programming) * Implement an organized search for reflections in the array data (optimization) * Use these condensed results to infer the geometry of the space (visualization of results)

Requirements

* Basic knowledge of sampling theory * Basic MATLAB programming knowledge. * Curiosity to learn new things. * Ability to autonomously identify problems and work on their solution.