Multipitch estimation and instrument recognition by exemplar-based sparse representation

Ikuo Degawa, Kei Sato, Masaaki Ikehara

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper investigates the pitch estimation and the instrument recognition of music signals. A note exemplar is a spectrum segment of notes of the specific pitch and instrument, which is stored as a form of dictionary preliminarily. We describe the method of reconstructing a frame of musical signals as the linear combination of exemplars from the large exemplar dictionary with sparse (l1 minimized) coefficient vector. Reconstruction constraints are imposed to KL divergence of spectra, which is found to produce better results than Euclidean distance. The proposed algorithm shows the ability to transcript music pieces with relatively many notes per a frame and to divide the instrument explicitly through some experiments.

Original languageEnglish
Title of host publicationConference Record - Asilomar Conference on Signals, Systems and Computers
PublisherIEEE Computer Society
Pages560-564
Number of pages5
ISBN (Print)9781479923908
DOIs
Publication statusPublished - 2013
Event2013 47th Asilomar Conference on Signals, Systems and Computers - Pacific Grove, CA, United States
Duration: 2013 Nov 32013 Nov 6

Other

Other2013 47th Asilomar Conference on Signals, Systems and Computers
CountryUnited States
CityPacific Grove, CA
Period13/11/313/11/6

Fingerprint

Glossaries
Experiments

Keywords

  • instrument recognition
  • l1 regularized minimization
  • note exemplar
  • pitch estimation

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Signal Processing

Cite this

Degawa, I., Sato, K., & Ikehara, M. (2013). Multipitch estimation and instrument recognition by exemplar-based sparse representation. In Conference Record - Asilomar Conference on Signals, Systems and Computers (pp. 560-564). [6810341] IEEE Computer Society. https://doi.org/10.1109/ACSSC.2013.6810341

Multipitch estimation and instrument recognition by exemplar-based sparse representation. / Degawa, Ikuo; Sato, Kei; Ikehara, Masaaki.

Conference Record - Asilomar Conference on Signals, Systems and Computers. IEEE Computer Society, 2013. p. 560-564 6810341.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Degawa, I, Sato, K & Ikehara, M 2013, Multipitch estimation and instrument recognition by exemplar-based sparse representation. in Conference Record - Asilomar Conference on Signals, Systems and Computers., 6810341, IEEE Computer Society, pp. 560-564, 2013 47th Asilomar Conference on Signals, Systems and Computers, Pacific Grove, CA, United States, 13/11/3. https://doi.org/10.1109/ACSSC.2013.6810341
Degawa I, Sato K, Ikehara M. Multipitch estimation and instrument recognition by exemplar-based sparse representation. In Conference Record - Asilomar Conference on Signals, Systems and Computers. IEEE Computer Society. 2013. p. 560-564. 6810341 https://doi.org/10.1109/ACSSC.2013.6810341
Degawa, Ikuo ; Sato, Kei ; Ikehara, Masaaki. / Multipitch estimation and instrument recognition by exemplar-based sparse representation. Conference Record - Asilomar Conference on Signals, Systems and Computers. IEEE Computer Society, 2013. pp. 560-564
@inproceedings{81821f5e76fc4f5396cac0a3533de581,
title = "Multipitch estimation and instrument recognition by exemplar-based sparse representation",
abstract = "This paper investigates the pitch estimation and the instrument recognition of music signals. A note exemplar is a spectrum segment of notes of the specific pitch and instrument, which is stored as a form of dictionary preliminarily. We describe the method of reconstructing a frame of musical signals as the linear combination of exemplars from the large exemplar dictionary with sparse (l1 minimized) coefficient vector. Reconstruction constraints are imposed to KL divergence of spectra, which is found to produce better results than Euclidean distance. The proposed algorithm shows the ability to transcript music pieces with relatively many notes per a frame and to divide the instrument explicitly through some experiments.",
keywords = "instrument recognition, l1 regularized minimization, note exemplar, pitch estimation",
author = "Ikuo Degawa and Kei Sato and Masaaki Ikehara",
year = "2013",
doi = "10.1109/ACSSC.2013.6810341",
language = "English",
isbn = "9781479923908",
pages = "560--564",
booktitle = "Conference Record - Asilomar Conference on Signals, Systems and Computers",
publisher = "IEEE Computer Society",

}

TY - GEN

T1 - Multipitch estimation and instrument recognition by exemplar-based sparse representation

AU - Degawa, Ikuo

AU - Sato, Kei

AU - Ikehara, Masaaki

PY - 2013

Y1 - 2013

N2 - This paper investigates the pitch estimation and the instrument recognition of music signals. A note exemplar is a spectrum segment of notes of the specific pitch and instrument, which is stored as a form of dictionary preliminarily. We describe the method of reconstructing a frame of musical signals as the linear combination of exemplars from the large exemplar dictionary with sparse (l1 minimized) coefficient vector. Reconstruction constraints are imposed to KL divergence of spectra, which is found to produce better results than Euclidean distance. The proposed algorithm shows the ability to transcript music pieces with relatively many notes per a frame and to divide the instrument explicitly through some experiments.

AB - This paper investigates the pitch estimation and the instrument recognition of music signals. A note exemplar is a spectrum segment of notes of the specific pitch and instrument, which is stored as a form of dictionary preliminarily. We describe the method of reconstructing a frame of musical signals as the linear combination of exemplars from the large exemplar dictionary with sparse (l1 minimized) coefficient vector. Reconstruction constraints are imposed to KL divergence of spectra, which is found to produce better results than Euclidean distance. The proposed algorithm shows the ability to transcript music pieces with relatively many notes per a frame and to divide the instrument explicitly through some experiments.

KW - instrument recognition

KW - l1 regularized minimization

KW - note exemplar

KW - pitch estimation

UR - http://www.scopus.com/inward/record.url?scp=84901247544&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84901247544&partnerID=8YFLogxK

U2 - 10.1109/ACSSC.2013.6810341

DO - 10.1109/ACSSC.2013.6810341

M3 - Conference contribution

AN - SCOPUS:84901247544

SN - 9781479923908

SP - 560

EP - 564

BT - Conference Record - Asilomar Conference on Signals, Systems and Computers

PB - IEEE Computer Society

ER -