Use of Vocal Source Features in Speaker Segmentation | IEEE Conference Publication | IEEE Xplore