Automatic noise source recognition

Didier Dufournet, Philippe Jouenne, Adam Rozwadowski
1998 Journal of the Acoustical Society of America  
Madras is an intelligent noise monitoring system able to classify the acoustic signatures of different kinds of noise sources. The purpose of this paper is to present the use and the adaptation of several methods of signal processing and artificial intelligence to acoustic signal characterization and specially the "meta-class" recognition step. The MADRAS system works with Symphonie, a PC-based dual-channel real time analyzer developed by 01dB.
doi:10.1121/1.422237 fatcat:ghpw52oy5rg3xli3b7f4n2quuu