Distributed environmental sound monitoring and auditory perception

Distributed environmental sound monitoring systems are increasingly being applied for assessing noise pollution in urban context. As microphones and computing devices become cheaper and more reliable, networks can be vastly expanded and individual measurement nodes can perform more tasks than just logging sound levels. Consequently, the operator is faced with the problem of how to analyze the large amount of data that is generated by such a network. One of the research lines at the acoustics group of Ghent University is to develop new algorithms for extracting useful information from acoustic measurement networks. A human-centered approach is followed, in which algorithms that closely mimic human perception are preferred. Topics that will be covered in this presentation are the detection of salient events that could attract human attention, the construction of an acoustical summary, the estimation of sound source direction and sound source recognition.Distributed environmental sound monitoring systems are increasingly being applied for assessing noise pollution in urban context. As microphones and computing devices become cheaper and more reliable, networks can be vastly expanded and individual measurement nodes can perform more tasks than just logging sound levels. Consequently, the operator is faced with the problem of how to analyze the large amount of data that is generated by such a network. One of the research lines at the acoustics group of Ghent University is to develop new algorithms for extracting useful information from acoustic measurement networks. A human-centered approach is followed, in which algorithms that closely mimic human perception are preferred. Topics that will be covered in this presentation are the detection of salient events that could attract human attention, the construction of an acoustical summary, the estimation of sound source direction and sound source recognition.C