Primi speech recognition based on deep neural network
Wenjun Hu, Meijun Fu, Wenlin Pan · 2016
In order to improve the performance of Primi speech recognition system, a novel method based on deep neural network has been proposed. The deep neural network has two distinct characteristics, one is a high-capacity, and the other is a highly complex network structure. On the Kaldi platform, the neural network, containing four hidden layers, which used to deal with the Primi speech recognition. The results of Test 1 showed that: the Word Error Rate in Primi speech recognition was reduced by 47.9%, 4.2% and 1.7% respectively using the deep neural network with four tanh hidden layers compared to that using the mono-phone model based on GMM-HMM, optimized-tri-phone model based on GMM-HMM and optimized-SGMM model. So, the deep neural network could not only complete large vocabulary speech recognition, but also implemented its recognition rate significantly higher than the traditional HMM.