摘要
随着国家对教育的重视,中国国民平均受教育水平不断提高,越来越多的家庭意识到早期教育的重要性。为了解决现存的城乡之间早期教育资源差距大和人们对早期教育App(application)信赖度不高的问题,设计了基于卷积神经网络(convolutional neural networks,CNN)深度学习技术的早期教育系统。系统将CNN技术运用于图像识别和儿童语音识别;内置人工智能助手小铃同学来引导儿童对事物的探索;设置了家长端,通过采集儿童端的儿童使用数据,系统将其制成图表的形式反馈给家长。本系统将语音识别和CNN技术相结合时,着重研究儿童声音的特点并以此为基础进行声音模型训练,提高了识别的精准度;通过数据挖掘分析技术,在家长端生成儿童的使用报告并获得基于此报告的儿童兴趣方向推荐和相关课程信息,为家长提供了了解儿童兴趣方向的途径。
As the country attaches importance to education,the average education level of Chinese citizens continues to increase,and more and more families are aware of the importance of early education.In order to solve the existing problems of the large gap in early education resources between urban and rural areas and people s low trust in early education apps(applications),an early education system was designed on the basis of convolutional neural networks(CNN)deep learning technology.The system used CNN technology for image recognition and children s speech recognition;it had built-in artificial intelligence assistant Xiaoling to guide children s exploration of things;it was equipped with a parent terminal,to which the system fed back in the form of a chart the children s usage data collected from the children terminal.When the system combined speech recognition and CNN technology,it focused on the characteristics of children s voice and conducted voice model training based on this,which improved the accuracy of recognition.Through data mining and analysis,the system generated a children s usage report on the parent terminal and obtained children s interest direction recommendations and relevant course information based on this report,providing parents with a way to understand the direction of children s interests.
作者
沈梦婷
岑岗
周闻
朱润锴
梁娅勤
SHEN Mengting;CEN Gang;ZHOU Wen;ZHU Runkai;LIANG Yaqin(School of Information and Electronic Engineering,Zhejiang University of Science and Technology,Hangzhou 310023,Zhejiang,China;School of Mechanical and Energy Engineering,Zhejiang University of Science and Technology,Hangzhou 310023,Zhejiang,China)
出处
《浙江科技学院学报》
CAS
2020年第6期590-594,共5页
Journal of Zhejiang University of Science and Technology
基金
国家级大学生创新创业训练计划项目(201911057011)。
关键词
卷积神经网络
深度学习
人工智能
早期教育
系统设计
convolution neural network(CNN)
deep learning
artificial intelligence
early education
system design
作者简介
通信作者:岑岗(1959-),男,浙江省象山人,教授,主要从事教育信息科学与技术研究。E-mail:gcen@163.com。
