Evaluation of Mandarin Pronunciation Based on Phonemic Weights and HMM Modeling
Ganqiao Dan · 2023
In this study, we constructed a Mandarin pronunciation evaluation system based on phoneme weights and HMM model, aiming to improve the accuracy and personalized adaptability of pronunciation evaluation. The experimental results show that the system performs well in diverse speech contexts and effectively supports personalized pronunciation learning. Future work will focus on improving the real-time performance of the system, extending the application scenarios, and further optimizing the user experience. This research provides new ideas and useful references for the application of speech processing technology in the field of language learning.