Design of English Speech Comprehensive Training System for Smart Phone

Smartphone software has the characteristics of abundant resources and simple operation, and has become an important tool for foreign language learning. This paper studies the key technologies of the English phonetic comprehensive training system for smartphones. First, text-to-speech conversion, combined with knowledge of linguistics and psychology, with the support of computer and other hardware environments, converts text information into natural speech streams. Second, the prosody generation based on neural network can spontaneously master prosody rules through simulation learning, and establish data association between prosody in speech and language, so as to meet the requirements of prosody processing. Third, speech recognition based on neural network is based on data preprocessing and speech feature extraction, and more accurate data can be obtained by using bidirectional recurrent neural network.