Pattern Recognition Based Music Style Recognition and Teaching Application in Higher Education Music Education

This paper presents a summary of a range of characteristic parameters that define the tone features. This is achieved by studying the time-frequency and frequency characteristics of music signals with different instrumental timbres, and it represents the characteristics of the music in various frequency bands and time domains. The optimized DTW pattern recognition algorithm achieves the classification of music styles. The conducted experiments clearly recognized several basic violin bowing styles. Jazz’s classification and recognition effect is 80% accurate. The accuracy rate of the music brief spectrum recognition exceeded 95%. The teaching method based on music pattern recognition has a significant teaching effect in the knowledge and skill dimensions, with a Sig. value of 0.001.