面向抒情歌曲旋律的钢琴自动伴奏算法  被引量:2

Automatic Piano Accompaniment Algorithm for the Melodies of Lyric Songs

在线阅读下载全文

作  者:曹西征[1,2] 牛靖雯 秦杰[1] 孙志勇[1] 

机构地区:[1]河南师范大学计算机与信息工程学院,河南新乡453007 [2]河南师范大学"智慧商务与物联网技术"河南省工程实验室,河南新乡453007

出  处:《河南师范大学学报(自然科学版)》2016年第4期137-142,共6页Journal of Henan Normal University(Natural Science Edition)

基  金:国家自然科学基金(61402153);河南省重点科技攻关项目(122102210054);河南省高校青年骨干教师项目(2011GGJS-061)

摘  要:目前钢琴自动伴奏中存在着一定的不和谐问题,其中一个重要原因是由于和弦处理的固定性往往不能满足旋律的不确定性要求.为此,提出一种面向拍号固定的抒情歌曲旋律的自适应伴奏算法.首先,根据拍号对旋律进行小节分割和乐句整合,在传统三度关系的三和弦基础上定义更泛化的三元和弦,自适应获取三元和弦的内部音程关系;然后,自适应计算旋律音对和弦的贡献分值,以获取小节中所有可能的三元和弦;自动选取和弦个数,将选出的和弦与知识库进行匹配,将匹配成功的和弦添加合适的伴奏音型,最终生成伴奏.实验结果表明,人们不容易将该算法生成的伴奏与人工的伴奏进行区分.At present, the disharmonious problem exists in the automatic piano accompaniment. One of the important reasons is that the stationarity of chord processing often canr t meet the requirement of the uncertainty of melody. Therefore, this paper proposes an adaptive accompaniment algorithm for the melodies of lyric songs whose time signature is fixed. First, the segmentation of bars and the integration of phrases are conducted in melody based on time signature. Second, the more gen- eralized ternary chords are defined on the basis of the traditional triads of three intervals, and the inner intervals of the ternary chords are obtained adaptively. Third, the melody tone's contributions to the choi'd are adaptively computed to get all the possi- ble ternary chords in bars. Fourth, the number of chords is obtained automatically, and some of them are matched with their knowledge base. Finally, the chords which are matched successfully are added a suitable accompaniment figure, and the piano accompaniments are generated. The experiment results show that people can't easily distinguish the accompaniments generated by this algorithm from that generated by manual.

关 键 词:自动伴奏 抒情歌曲 自适应伴奏 三元和弦 

分 类 号:TP391[自动化与计算机技术—计算机应用技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象