Combination of a hidden tag model and a traditional n-gram model: a case study in czech speech recognition.

Pavel Krbec, Petr Podvesk, Jan Hajic

Browse the full Interspeech paper archive.