Microphone Array Based Surveillance Audio Classification

Abstract

The work assessed seven classifiers and two beamforming algorithms for detecting surveillance sound events. The tests included the use of AWGN with -10 dB to 30 dB SNR and Data Augmentation (DA). The results showed that the combination of Support Vector Machine (SVM) and Delay-and-Sum (DaS) scored the best accuracy (up to 86.0%), but had high computational cost (≈ 79 ms), mainly due to DaS and DA. The use of Stochastic Gradient Descent (SGD) also seems to be a good alternative since it has achieved good accuracy either (up to 85.3%), but with quicker processing time (≈ 25 ms).