In this paper, a deep learning approach for classification of cough sound segments is presented. The architecture of the network is based on a pre-trained network and the spectrogram images of three recording channels have been extracted for the sake of training the network. The classification accuracy based on three recording channels is 92% for a binary classification model and the network converges fast. Two classification models based on binary and multi-class problems are proposed. Relevant classification parameters including the Receiver Operating Characteristic (ROC) curve are reported.
© 2020 by Walter de Gruyter Berlin/Boston
This work is licensed under the Creative Commons Attribution 4.0 International License.