Článek prezentuje implementaci automatického rozpoznávače řeči v jazyce Java. Systém podporuje akustické modely založené na neuronových sítích nebo na Markovových modelech se spojitou distribuční funkcí a dokáže pracovat s ůmodely softwaru HTK. (cs)
We present an implementation of our automatic speech recognizer in Java programming language. The system supports neural network based acoustic models as well as more traditional continuous densit hidden Markov models and can read HTK trained models.
We present an implementation of our automatic speech recognizer in Java programming language. The system supports neural network based acoustic models as well as more traditional continuous densit hidden Markov models and can read HTK trained models. (en)