Classification of Snoring Sounds

Kılıç R., Kumbasar N., Sağlam H. K., Oral E. A., Özbek İ. Y.

ISASE 2021, Erzurum, Türkiye, 7 - 09 Nisan 2021, ss.387-390

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Basıldığı Şehir: Erzurum
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.387-390
  • Atatürk Üniversitesi Adresli: Evet

Özet

— Snoring is a disturbance in breathing caused by air flow caused by tissue vibration in different parts of the upper respiratory tract during sleep. It can be a symptom of a serious illness such as sleep apnea. For this reason, it is of great importance that the origin of snoring sounds is determined and treated correctly. In this study, the MPSSC data set containing snoring sounds was used. The data set contains 828 audio files taken from 224 subjects. In the study, features were obtained with Mel-Frequency Cepstral Coefficients (MFCC), Pitch, SCAT and Gaussian Mixture Model (GMM) to determine the location of snoring sounds in the upper respiratory tract. Classification was made with K-Nearest Neighbors (KNN) and Support Vector Machine (SVM). In the classification of snoring sounds, 57.87% success rate was achieved with the Scat + GMM + SVM model as an initial study.