Automatic classification of environmental noise events by hidden Markov models

The automatic classification of environmental noise sources from their acoustic signatures recorded at the microphone of a noise monitoring system (NMS) is an active subject of research nowadays. This paper shows how hidden Markov models (HMMs) can be used to build an environmental noise recognition system based on a time-frequency analysis of the noise signal. The performance of the proposed HMM-based approach is evaluated experimentally for the classification of five types of noise events (car, truck, moped, aircraft, train). The HMM-based approach is found to outperform previously proposed classifiers based on the average spectrum of the noise event with more than 95% of correct classifications. For comparison, a classification test is performed with human listeners for the same data which shows that the best HMM-based classifier outperforms the "average" human listener who achieves only 91.8% of correct classification for the same task.