On comparing and combining intra-speaker variability compensation and unsupervised model adaptation in speaker verification | Publicación