Use of VTL-wise models in feature-mapping framework to achieve performance of multiple-background models in speaker verification.

Achintya Kumar Sarkar, Srinivasan Umesh

Browse the full ICASSP paper archive.