Recognition of Sounds Observed in Daily Living Activity of Elderly People

Abstract

An automated system is proposed to recognize the distressed situation form the daily living activity of human sounds. Elderly people are living as dependents in various countries but somehow this trends is much deviated now and people started living on their own. Due to illness or other medical occupations they prefer to stay at home rather than doing an office based job. The automated system will help in raising the living standard of such individuals along with those who are caretakers and cannot monitor them all the time. The sound activity detection model is proposed which captures and recognizes the sound of the entire day activity of an individual. Three Benchmark datasets are used to test our proposed model. The dataset used for our system are Real World Computing Partnership Sound Database in Real Acoustical Environment (RWCP-DB), Urban Sound8K and ESC10 data set. We used Linear Spectrogram, MFCC, Gamma tone Spectrogram as a base line for feature extraction using CNN. We proposed two models based on CNN and CNN-SVM architecture and also trained Alex Net and Goggle Net using transfer learning. Our system performed well on different combinations of features and showed improved classification accuracy. Our system performed well in comparison with the other methods reported in literature.