H- and C-level WFST-based large vocabulary continuous speech recognition on Graphics Processing Units.

Jungsuk Kim, Kisun You, Wonyong Sung

Browse the full ICASSP paper archive.