基于卷积神经网络的标准课堂教学语音情感优化的方法研究  

Research on the Method of Optimizing Speech Emotions in Standardized Classroom Teaching Based on Convolutional Neural Networks

作　　者：邓帅 吴筝 DENG Shuai;WU Zheng(Faculty of Education,Tianjin Normal University)

机构地区：[1]天津师范大学教育学部

出　　处：《中国标准化》2024年第8期241-244,共4页China Standardization

摘　　要：在人工智能时代下,深度学习技术逐渐融入教育研究领域的各个方面。教师教学语言是课堂教学的主要方式,教学情感作为教学评价的主要评估方式,深刻影响着教师的教学效果。本文基于卷积神经网络模型对教师语音情感进行识别,以语音情感描述模型和情感教学理论作为理论基础,按照“数据库建立—模型搭建—实践应用”的研究路径开展教师标准课堂教学语音情感的研究,建立教师课堂语音数据库,构建教师语音情感评价量表,还原真实课堂的精准采集、助力教师评价的高效开展,以此优化教师教学语音情感,赋能教学改进。In the era of artifi cial intelligence,deep learning technology is gradually integrating into various aspects of educational research.Teacher’s teaching language is the main method of classroom teaching,and teaching emotions,as the main evaluation method of teaching evaluation,deeply affect the teaching effectiveness of teachers.Based on the convolutional neural network model,teachers’speech emotion recognition is carried out.Using the speech emotion description model and emotional teaching theory as the theoretical foundation,the research path of“database establishment—model construction—practical application”is followed to carry out the research on teachers’standardized classroom teaching speech emotions.A teacher’s classroom speech database is established,and a teacher’s speech emotion evaluation scale is constructed to restore accurate collection of real classrooms and assist in the effi cient implementation of teacher evaluation,by optimizing teachers’teaching speech emotions and empowering teaching improvement.

关 键 词：教师语音情感 卷积神经网络 教学改进 深度学习 

分 类 号：G424[文化科学—课程与教学论]