Automatic Piano Accompaniment Generation Method by Drum Rhythm Features with Selectable Difficulty Level

Kazuma Komiya,Ryotaro Okada, Ayako Minematsu,Takafumi Nakanishi

Studies in computational intelligence(2023)

引用 0|浏览1
暂无评分
摘要
This paper shows the automatic piano accompaniment generation method by drum rhythm features with selectable difficulty levels. In general, when arranging a piece of music to be played on the piano by oneself, one needs to have an exceptionally high level of knowledge regarding the arrangement of the left-hand part of that accompaniment. This paper presents a new system that takes as input the URL of a piece of music uploaded on a video distribution site and outputs several accompaniment scores of different difficulty levels. This system obtains the drum sounds and chord progressions of a piece of music from the URL input by the user and generates an accompaniment score. This system focuses on the rhythm of the music. It generates an accompaniment score that reproduces the piece’s rhythm by selecting an accompaniment pattern that has a rhythm close to that of the music from among several accompaniment patterns prepared in advance. In addition, it generates scores with multiple difficulties by designing accompaniment patterns by difficulty level. It makes it possible to suggest musical scores that match the performer’s skill level. By implementing this system, piano players will have more opportunities to select and play music from a vast and diverse range of music based on their preferences.
更多
查看译文
关键词
drum rhythm features,piano,selectable difficulty level
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要