Combining the glottal mixture model (GLOMM) with UBM for speaker recognition | IEEE Conference Publication | IEEE Xplore