Next: Task 3.1: Status
Up: Task 3.1: Markov Model
Previous: Task 3.1: Markov Model
Markov model based grammars (e.g. trigrams) are the most commonly
used, consistent and effective language models used in LVCSR. It is
the objective of this task to provide efficient, scalable tools for
the use of such grammars, and to extend them via the use of higher-
(or variable-) order Markov models and class-based models.