Speaker adaptive training in deep neural networks using speaker dependent bottleneck features.

Rama Doddipatla

Browse the full ICASSP paper archive.