Snoring Event Detection Using Machine Learning Techniques

点击翻译按钮获取中文摘要

Snoring is one of the major symptoms of a sleep disorder. Surveys conducted by the National Sleep Foundation**(1999-2004) have revealed that at least 40 million Americans suffer from over 70 different sleep disorders.**Snoring seems underestimated because people cannot recognize the seriousness of their snoring since it occurs**during sleep. Manual recording and examination of a person's respiratory sounds for the entire night can be a**very time-consuming and operator-dependent task. Therefore, an automatic sound recording technique is**desirable. Furthermore, existing solutions in the market for detection of snoring are not designed to address the**core issue itself. Our aim is to both diagnose and address the core issue of snoring using our active antisnoring**solution in the users' bedroom. Almost all of the snore related products make physical contact with the snorer**which result in painful, uncomfortable or disruptive experiences. Smart Nora is the first contact-free, non-invasive**active product on the market to help stop snoring. Although the current product works fine with average**scenarios, the company receives a number of returns and negative feedbacks based on snore detection algorithm**of the current product. Customers often complain about the product as 'not catching the snore', 'activating too**much', or giving 'false alarms' when there is no snore. This project will investigate the development and**implementation of a machine learning algorithm which aims to distinguish the snore events from different type**background noises by using machine learning algorithms and advanced signal processing techniques.