Variational Autoencoders for Anomaly Detection in Respiratory Sounds

This paper proposes a weakly-supervised machine learning- based approach aiming at a tool to alert patients about possible res- piratory diseases. Various types of pathologies may affect the respira- tory system, potentially leading to severe diseases and, in certain cases, death. In general, effective prevention practices are considered as major actors towards the improvement of the patient’s health condition. The proposed method strives to realize an easily accessible tool for the auto- matic diagnosis of respiratory diseases. Specifically, the method leverages Variational Autoencoder architectures permitting the usage of train- ing pipelines of limited complexity and relatively small-sized datasets. Importantly, it offers an accuracy of 57%, which is in line with the exist- ing strongly-supervised approaches.