A Sound Source-based Intelligent Context Awareness System using CNN

A sound source-based intelligent context awareness system using convolutional neural network (CNN) is proposed and the sound recognition ratio was analyzes in this paper. The sound recognition rate was found to be proportional to the distance between the smartphone and the microphone. It is confirmed that the sound source is difficult to be detected when the distance between the microphone and the sound source is far greater than 1m on 1
^st
experiment. However when the distance between the sound source and the microphone is 1m, the performance of the sound recognition rate is 40%-55% on 2
^nd
experiment.