A system and method for detecting the presence of a particular source of audible content using an automatic speech recognizer component which is trained to recognize specific non-speech sounds. Training of the automatic speech recognizer component is accomplished by providing an appropriate model to be recognized and by tuning the automatic speech recognizer to the sounds created by the desired stimulus. The automatic speech recognizer can additionally be provided with processing means for distinguishing the desired stimulus from an irrelevant or unwanted stimulus which may have acoustic signatures with a high degree of similarity to that of the desired stimulus. The invention can be implemented in a portable detector to allow detection of a specific acoustic source at any location.