Machine models of music
Selected papers of the 9th annual ACM SIGACT-SIGMOD-SIGART symposium on Principles of database systems
Manipulation of music for melody matching
MULTIMEDIA '98 Proceedings of the sixth ACM international conference on Multimedia
Towards a digital library of popular music
Proceedings of the fourth ACM conference on Digital libraries
Melodic matching techniques for large music databases
MULTIMEDIA '99 Proceedings of the seventh ACM international conference on Multimedia (Part 1)
Principles of Database Systems
Principles of Database Systems
An Informal Introduction to Constraint Database Systems
CDB '97 Second International Workshop on Constraint Database Systems, Constraint Databases and Their Applications
Warping indexes with envelope transforms for query by humming
Proceedings of the 2003 ACM SIGMOD international conference on Management of data
Hi-index | 0.00 |
A more natural way of searching a song is by humming or singing, especially when people can not remember the title, singer or lyric, but can remember a piece of melody. This paper is about how to represent melody so that such query is possible and efficient. We propose a constraint-based method. Considering both local and global constraints of pitch and duration, we can derive a melody's canonical form by using linear algorithm Reduce and Weight. This representation is at least as precise as interval-based method, but avoid much of redundant searching since it is more specialized. The significance of our method is that the tonal context, a key factor of music, is taken into account. This feature makes it not only a good candidate for building an efficient music database system, but also a foundation for further investigation of music, such as style analysis, pattern extraction, and finding other constraints.