摘要
提出一种新型普通话测试系统评分机制——基于人工智能的评分机制(PSCAI).与现有的中值滤波评分机制不同,PSCAI是一种完全基于神经元连接的网络模型,它能不断地接收、存储各种语音信息,并把感觉足够强的语音模式记忆下来,这一过程更接近于人脑的学习、记忆过程.实验结果证明,PSCAI学习评判的效率高,在评判不用语音时不会影响已有的语音,同时具有很强的语音识别能力.
A new kind of mechanism,Mechanism of Putonghua Shuiping Ceshi based on Artificial Intelligence(PSCAI), is proposed in this paper.Different from the current median filter,PSCAI is completely based on the network model of neuron connection.It can receive and store sound information constantly,and memorize the sound information or patterns which result strong feeling.This process is quite similar to the learning and memorization process of human brain.Simulation results show that PSCAI performs well in sound recognition and can study with very high efficiency.The existing sound patterns are not influenced in judging useless sound.
出处
《玉溪师范学院学报》
2013年第12期21-24,共4页
Journal of Yuxi Normal University
