Application of Deep Learning in Urban Sounds Classification

Abstract

There is a large area of research for automatic sound classification with countless real world applications. Even though there is a large body of research in related audio fields such as speech and music, work on the classification of environmental sounds is comparatively scarce. Hence, when we observe the recent advancements in the field of image classification where convolutional neural networks are used to classify images with high accuracy and at scale, we come across questions on applicability of these techniques in other domains, such as sound classification, where discrete sounds happen over time. This paper is the analysis of the project that tests audio samples and classifies them accordingly using Deep Learning techniques, namely, Multi-Level Perceptron (MLP) and modified Convolutional Neural Networks (CNN).