Research on the two-way influence of music education and information technology integration on teaching effect and student participation

Abstract

An automatic transcription system can provide a symbolic representation of performance content, and in music education applications, it can assist teachers or students in recording the performance of a piece. This study examines the automatic transcription scenario for both audio and image input. First, a CQT transform algorithm is proposed, incorporating an energy equalization module to extract key features from the musical score and a convolutional neural network to complete the detection of the audio. Secondly, we utilize a graph convolutional neural network to detect visual movements of hands and keys. Finally, we conducted an empirical study to verify the impact of information technology on the effectiveness of music teaching and student engagement. By carrying out experimental simulation tests on data sets 1 and 2, it is found that not only the CQT algorithm can accurately respond to the digital characteristics of the spectrum, but also the system performance after energy equalization processing is better compared to standard CQT, with the F1 value in data set 1 improved to 95.57% and that in data set 2 improved to 95.55%. In addition, the visual transcription system recognized each finger of the right hand with greater accuracy than those of the left hand and detected black keys better than white keys. The percentage of individuals rated excellent grades increased by 15.1% before and after receiving information technology training, while the percentage of individuals rated poor grades decreased by 3.42%. According to the student’s classroom participation questionnaire, the average score for each dimension was above 3.

Keywords

• cqt transform
• energy equalization
• audio transcription
• visual transcription
• music instruction
• 97m80