Sound Classification Using Convolutional Neural Networks

Abstract

The performance of Sound Classification like Urban Sound Classification, Environmental Sound Classification, etc. has been significantly improved with the evolution of Deep Neural Networks (DNN). In this paper, we propose a model which uses Convolutional Neural Networks (CNN) for classification of sound based on the spectrograms obtained for different sound samples. The model can be used for Deforestation detection, Gunshot detection in urban areas and also for detecting unusual sounds in streets like a cry for help, tyres screeching etc. at odd hours.