Analysis of Audio Signals Using Deep Learning Algorithms Applied to COVID Diagnostic Systems

Abstract

In recent years the application of deep learning algorithms in the subdomain of audio analysis has grown rapidly, however it is a topic that can be complex for students and researchers who have a first approach and want to develop an application in this field. The use of deep learning techniques applied to audio signals has allowed the development of a wide variety of useful tools in our daily lives, from virtual assistants to medical applications. This article presents a literature review of the main techniques that have been used in recent years for analysis, feature extraction and classification from audio spectra or spectrograms, as well as examples of application in the context of the COVID-19 pandemic in which multiple related projects have emerged, such as diagnostic systems. The techniques addressed are recurrent neural networks (RNN), convolutional neural networks (CNN) and generative adversarial networks (GAN). It is intended that the reader will be able to acquire this knowledge from a simple perspective and that this information will be useful in their research or development.