Integrating multi-microphone devices and AI for noise event identification in unattended monitoring

This paper explores the use of multi-microphone devices and artificial intelligence (AI) for the identification of noise events in unattended noise monitoring. The system presented aids in the identification and localization of dominant sound sources using time difference of arrival techniques. By incorporating AI technology, the system's software automates the extraction of non-relevant sounds. The algorithms enable removal of undesired components, resulting in a more accurate measurement. This article discusses the benefits and limitations associated with employing automatic processes in noise monitoring and analysis software. Finally, a current application and prospects for future advancements in the field are presented.