College of Music and Dance, Zhengzhou Normal University, Zhengzhou 450044, China.
Comput Intell Neurosci. 2022 Sep 6;2022:4520953. doi: 10.1155/2022/4520953. eCollection 2022.
It is a challenge for the current music teaching system to carry out teaching according to the difference of score difficulty and realize automatic grading. Therefore, identifying the difficulty of music score according to pitch is the key to individualize music teaching resources. This paper summarizes and analyzes the problem of pitch feature extraction in music teaching. In the pitch extraction, the audio signal is divided into frames, and the feature matching of high-pitched content in music teaching resources is realized by smoothing the pitch sequence. In addition, the pitch feature extraction algorithm in MIDI music score files is proposed, and the pitch feature matching model is constructed. Finally, a music tutoring system based on pitch feature matching is designed, including a music score learning tool, overall structure of system, and interaction between teachers and students. Tutoring strategies include three main functions: learning suggestions of knowledge points, skills in practice and training, and learning path adjustment. This study is helpful to further improve the music teaching model and realize intelligent and personalized music learning.
当前的音乐教学系统在根据乐谱难度进行教学和实现自动评分方面存在挑战。因此，根据音高识别乐谱的难度是实现音乐教学资源个性化的关键。本文总结和分析了音乐教学中音高特征提取的问题。在音高提取中，音频信号被分为帧，通过平滑音高序列来实现音乐教学资源中音高内容的特征匹配。此外，还提出了 MIDI 乐谱文件中音高特征提取算法，并构建了音高特征匹配模型。最后，设计了一个基于音高特征匹配的音乐辅导系统，包括音乐乐谱学习工具、系统的总体结构以及师生之间的交互。辅导策略包括三个主要功能：知识点学习建议、实践和训练技能以及学习路径调整。本研究有助于进一步完善音乐教学模式，实现智能个性化音乐学习。
