2016 Fiscal Year Final Research Report
Development of music information retrieval system for instrumental music by using automatic music-to-onomatopoeias converter
Project/Area Number |
25330140
|
Research Category |
Grant-in-Aid for Scientific Research (C)
|
Allocation Type | Multi-year Fund |
Section | 一般 |
Research Field |
Multimedia database
|
Research Institution | Osaka Institute of Technology |
Principal Investigator |
|
Research Collaborator |
KAWASHIMA Kohei
KUBO Hayato
KUROKAWA Yutaro
KOMAI Riki
SATO Tomoya
SUGITA Yusuke
SUGIMOTO Yuta
HISAOKA Akimitsu
|
Project Period (FY) |
2013-04-01 – 2017-03-31
|
Keywords | 楽曲検索システム / 擬音語歌唱 / 音程抽出 / 歌唱音声認識 |
Outline of Final Research Achievements |
In this study, we have developed a music information retrieval (MIR) system for instrumental music, which uses singing voice by onomatopoeias as query. Relationship between music and its representation by onomatopoeias can be used for retrieval in addition to melody information. In order to realize such MIR system, following four methods has been developed; automatic music-to-onomatopoeias converter, high performance melody information extractor and lyrics recognizer from singing voice, and high speed retriever. From experimental results, the converter outputted appropriate onomatopoeias representations. The extractor and the recognizer outperformed conventional methods, and the retriever suppressed about 10% amount of calculation.
|
Free Research Field |
音声情報処理
|