Collegiate Collaborative Piano Innovation Education Model Empowered by Artificial Intelligence

Abstract

In the era of big data, the development of piano education and teaching has irreversibly shifted towards an intelligent piano education model. In order to optimize the effectiveness of piano teaching in colleges and universities, this paper proposes a new model of intelligent piano education for college and university cooperation based on the Bayesian net model of automatic piano composition and dynamic gesture recognition and correction of piano playing. Combined with the rules of piano music accompaniment composition, it outlines the method of playing the piano left and right hand. Propose various genres and features and analyze the detection rate of Bayesian net models for different genres. The Camshaft algorithm is used to track and record the features of piano playing, while the DTW algorithm is used to detect and recognize piano fingering. Combine the DTW algorithm for gesture recognition with two constraints, endpoint fixation, and endpoint relaxation, to further test the recognition advantage of the DTW algorithm for piano fingering. The feasibility of the cooperative piano education model for colleges and universities proposed in this paper can be verified by comparing the students’ comprehensive music literacy scores in traditional piano lessons and intelligent piano lessons. The Bayesian net model and DTW algorithm both have a detection rate of more than 90% for piano music genres. Compared to traditional piano lessons, intelligent piano lessons can improve 25.12% of piano knowledge and 31.75% of piano repertoire scores, with the intelligent college cooperative piano education model demonstrating significant results.

Keywords

• automatic piano score
• camshift algorithm
• dtw algorithm
• gesture features
• piano education model
• 00a35