Two-step unsupervised speaker adaptation based on speaker and gender recognition and HMM combination.

Petr Cerva, Jan Nouza, Jan Silovsk

Browse the full Interspeech paper archive.