Speech separation of a target speaker based on deep neural networks | IEEE Conference Publication | IEEE Xplore