Discriminatively Trained Context-Dependent Duration-Bigram Models for Korean Digit Recognition.

Daniel Willett, Franz Gerl, Raymond Brueckner

Browse the full ICASSP paper archive.