A Deep Learning Model to Snore Detection Using Smart Phone

Snoring is a harsh and irritating sound in our throats as wind streams past loosen tissues, allowing the tissues to vibrate when we inhale. Once in a while, nearly everybody snores, enabling the tissues to vibrate as we breathe. Although everybody snores, sometimes it could very well be an ongoing problem for particular people. It may even indicate an actual medical condition now and then. Helpless rest will cause colossal physical, emotional, and monetary problems caused by snoring. Resting as an afterthought rather than dozing on either the back is a simple and popular solution to snoring. In this paper, a deep learning model for snore detection and implementation of the software on an Android smartphone is discussed. An Android app is produced to decide if the audio sample is snore/non-snore using the real-time audio piece captured using the unit's internal microphone. It also operates on any phone-connected audio input unit. The Snore Detection deep learning model has achieved an accuracy of 98%. A prototype app consisting of the input device (microphone), Android Smart Phone (for Android app deployment) have been developed and tested successfully.