Research on machine detection of hearing recognition ability of hearing impaired children based on GMM
XU Jie1， HAN Xueqing2， LIAO Qingzhou3， LIAO Shengbin2
(1.School of Economics and Business Administration， Central China Normal University， Wuhan 430079， China；2.National Engineering Research Center for E-Learning， Central China Normal University， Wuhan 430079， China；3.Wuhan Vocational College of Software and Engineering， Wuhan 430205， China)
Abstract：The detection of hearing recognition ability of hearing impaired children mainly adopts the Mandarin version of Lin's seven-tone manual detection method in China， which has problems such as low speed and time-consuming. In response to these problems， a method for machine recognition of the Mandarin version of Lin's seven-tones is proposed to detect hearing impaired children's auditory recognition ability in this paper. A Gaussian Mixture Model is presented， which was trained using the constructed Lin's seven-tone dataset. The comparison experiments with the Hidden Markov Model were provided. And three indicators， i.e.， precision， recall and accuracy are used to evaluate these two models. The experimental results show that the Gaussian mixture algorithm outperforms the hidden Markov algorithm in three indicators， and can better realize the recognition of the Mandarin version of Lin's seven-tone test.
徐 杰,韩雪晴,廖庆洲,廖盛斌. 基于GMM的听障儿童听觉辨识能力机器检测研究[J]. 华中师范大学学报(自然科学版), 2023, 57(6): 807-812.
XU Jie,HAN Xueqing,LIAO Qingzhou,LIAO Shengbin. Research on machine detection of hearing recognition ability of hearing impaired children based on GMM. journal1, 2023, 57(6): 807-812.
2023, Vol. 57 
