RETRACTED: Research on automatic evaluation method of Mandarin Chinese pronunciation based on 5G network and FPGA

Abstract

In the automatic evaluation system, need to learn the standard mandarin of scoring method for teaching in native Chinese pronunciation. The most pronounced goal protocols focus on the context in which native speakers are unnatural. The new Hidden Markov Model (HMM) algorithm based on the traditional algorithm likely algorithm for Chinese syllables, whose final initial period is found in the area where evidence for the measurement of weight control has been found. Experiments have also shown that this algorithm is more effective than the traditional posterior recording algorithm of the Mandarin learning method. Force Hidden Markov Model- HMM Align alignment identification for each syllable and associated recording probability for speech evaluation via race-based reliability system applications. These processes could then be formalized as a linear combination after the overall assessment functions: phonics, tone, intensity, and rhythm. Because both linear and non-linear parameters are involved in the overall evaluation functions. Incorporates variation in pronunciation to generate structure through a novel approach that incorporates tons of sub-tones that represent the missing automatic sound models. The word level assessment achieved through the pronunciation is similar to that which in the future showed the singing ability being realized by the evaluation system in full-length pronunciation as a method.