Introduction.

The term ‘acoustic scene analysis' (ASA) describes the task of extracting information contained in an acoustic wavefield, such as the waveform itself or parameter describing the source of the wavefield. Since acoustic wavefields are processes spread-out in space and time it follows quite naturally that ASA is predominantly performed by evaluating the signals captured by a number of spatially distinct microphones, i.e. microphone arrays. A standard and widely applied vehicle for evaluating the microphone array signals is built upon classical array signal processing techniques [JD93, Tre02]. In this context, the term ‘classical' is used to denote signal processing algorithms, to be introduced below, that are applied directly to the individual microphones comprising the array. In contrast, the algorithms to be derived in this book are applied to signals that are obtained by transforming the microphone signals into a domain defined by the eigen-solutions of the acoustic wave equation in two- and three spatial dimensions. [ABSTRACT FROM AUTHOR]