The Hybrid Model of Speech Recognition Based on PNN and HMM

摘要

Hidden Markov model (HMM) and improved probabilistic neural network (PNN) model are usually used for speech recognition, there are respective superiorities of them and some limitations in particular condition. In this paper, by analyzing and researching on the above two models, we propose a kind of hybrid model of speech recognition and give an algorithm that adapt to the hybrid model. The results show that this model has better recognition effect than HMM.