Microphone array processing techniques for classroom-based videoconferencing

Abstract

This thesis is concerned with the design and development of microphone-array-processing techniques for videoconferencing applications in classroom environments. We argue that, in such environments, it is advantageous in terms of both performance and practical implementation to use widely distributed arrays. A likely consequence of the use of such arrays is th at the relative locations of the microphones will be unknown, however, most previously published microphone-array-processing techniques assume and require knowledge of the microphone locations. We propose two novel algorithms which are designed for use with widely distributed arrays of unknown geometry.