当前位置:主页 > 科技论文 > 信息工程论文 >

流行音乐主旋律提取技术综述

发布时间:2018-04-27 09:06

  本文选题:主旋律提取 + 多音音乐 ; 参考:《计算机科学》2017年05期


【摘要】:旋律是最重要的音乐要素之一,多应用于音乐内容分析、音乐创作、音乐教育、抄袭检测等方面。主旋律提取旨在从一段音乐中自动估计对应于主旋律单音音符序列的音高或基频。流行音乐一般属于复杂的多音音乐,因此主旋律提取面临着许多挑战。综述了主旋律提取的研究背景,将主旋律提取的典型方法分别从基于音高显著度、基于声源分离以及其他方法3个类别进行了阐述,最后介绍了主旋律提取的评价指标以及研究进展。
[Abstract]:Melody is one of the most important elements of music, which is used in music content analysis, music creation, music education, plagiarism detection and so on. The purpose of the main melody extraction is to automatically estimate the pitch or fundamental frequency corresponding to the sequence of monophonic notes from a piece of music. Popular music generally belongs to the complex polyphonic music, so the main melody extraction faces many challenges. This paper summarizes the research background of the main melody extraction, and describes the typical methods of the main melody extraction from three categories, which are based on pitch salience, sound source separation and other methods, respectively. Finally, the evaluation index and research progress of theme extraction are introduced.
【作者单位】: 复旦大学计算机科学技术学院;复旦大学上海市智能信息处理重点实验室;
【基金】:国家自然科学基金NSFC(61671156)资助
【分类号】:TN912.3


本文编号:1810105

资料下载
论文发表

本文链接:https://www.wllwen.com/kejilunwen/xinxigongchenglunwen/1810105.html


Copyright(c)文论论文网All Rights Reserved | 网站地图 |

版权申明:资料由用户8b1d8***提供,本站仅收录摘要或目录,作者需要删除请E-mail邮箱bigeng88@qq.com