Supervised nonnegative matrix factorization with Dual-Itakura-Saito and Kullback-Leibler divergences for music transcription

Hideaki Kagami; Masahiro Yukawa

doi:10.1109/EUSIPCO.2016.7760426

Supervised nonnegative matrix factorization with Dual-Itakura-Saito and Kullback-Leibler divergences for music transcription

Hideaki Kagami, Masahiro Yukawa

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

5 Citations (Scopus)

Abstract

In this paper, we present a convex-analytic approach to supervised nonnegative matrix factorization (SNMF) based on the Dual-Itakura-Saito (Dual-IS) and Kullback-Leibler (KL) divergences for music transcription. The Dual-IS and KL divergences define convex fidelity functions, whereas the IS divergence defines a nonconvex one. The SNMF problem is formulated as minimizing the divergence-based fidelity function penalized by the ℓ₁ and row-block ℓ₁ norms subject to the nonnegativity constraint. Simulation results show that (i) the use of the Dual-IS and KL divergences yields better performance than the squared Euclidean distance and that (ii) the use of the Dual-IS divergence prevents from false alarms efficiently.

Original language	English
Title of host publication	2016 24th European Signal Processing Conference, EUSIPCO 2016
Publisher	European Signal Processing Conference, EUSIPCO
Pages	1138-1142
Number of pages	5
Volume	2016-November
ISBN (Electronic)	9780992862657
DOIs	https://doi.org/10.1109/EUSIPCO.2016.7760426
Publication status	Published - 2016 Nov 28
Event	24th European Signal Processing Conference, EUSIPCO 2016 - Budapest, Hungary Duration: 2016 Aug 28 → 2016 Sept 2

Other

Other	24th European Signal Processing Conference, EUSIPCO 2016
Country/Territory	Hungary
City	Budapest
Period	16/8/28 → 16/9/2

ASJC Scopus subject areas

Signal Processing
Electrical and Electronic Engineering

Access to Document

10.1109/EUSIPCO.2016.7760426

Cite this

Supervised nonnegative matrix factorization with Dual-Itakura-Saito and Kullback-Leibler divergences for music transcription. / Kagami, Hideaki; Yukawa, Masahiro.
2016 24th European Signal Processing Conference, EUSIPCO 2016. Vol. 2016-November European Signal Processing Conference, EUSIPCO, 2016. p. 1138-1142 7760426.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Kagami, H & Yukawa, M 2016, Supervised nonnegative matrix factorization with Dual-Itakura-Saito and Kullback-Leibler divergences for music transcription. in 2016 24th European Signal Processing Conference, EUSIPCO 2016. vol. 2016-November, 7760426, European Signal Processing Conference, EUSIPCO, pp. 1138-1142, 24th European Signal Processing Conference, EUSIPCO 2016, Budapest, Hungary, 16/8/28. https://doi.org/10.1109/EUSIPCO.2016.7760426

@inproceedings{f7bdd76581f142b2895fb018d0a727cc,

title = "Supervised nonnegative matrix factorization with Dual-Itakura-Saito and Kullback-Leibler divergences for music transcription",

abstract = "In this paper, we present a convex-analytic approach to supervised nonnegative matrix factorization (SNMF) based on the Dual-Itakura-Saito (Dual-IS) and Kullback-Leibler (KL) divergences for music transcription. The Dual-IS and KL divergences define convex fidelity functions, whereas the IS divergence defines a nonconvex one. The SNMF problem is formulated as minimizing the divergence-based fidelity function penalized by the ℓ1 and row-block ℓ1 norms subject to the nonnegativity constraint. Simulation results show that (i) the use of the Dual-IS and KL divergences yields better performance than the squared Euclidean distance and that (ii) the use of the Dual-IS divergence prevents from false alarms efficiently.",

author = "Hideaki Kagami and Masahiro Yukawa",

year = "2016",

month = nov,

day = "28",

doi = "10.1109/EUSIPCO.2016.7760426",

language = "English",

volume = "2016-November",

pages = "1138--1142",

booktitle = "2016 24th European Signal Processing Conference, EUSIPCO 2016",

publisher = "European Signal Processing Conference, EUSIPCO",

note = "24th European Signal Processing Conference, EUSIPCO 2016 ; Conference date: 28-08-2016 Through 02-09-2016",

}

TY - GEN

T1 - Supervised nonnegative matrix factorization with Dual-Itakura-Saito and Kullback-Leibler divergences for music transcription

AU - Kagami, Hideaki

AU - Yukawa, Masahiro

PY - 2016/11/28

Y1 - 2016/11/28

N2 - In this paper, we present a convex-analytic approach to supervised nonnegative matrix factorization (SNMF) based on the Dual-Itakura-Saito (Dual-IS) and Kullback-Leibler (KL) divergences for music transcription. The Dual-IS and KL divergences define convex fidelity functions, whereas the IS divergence defines a nonconvex one. The SNMF problem is formulated as minimizing the divergence-based fidelity function penalized by the ℓ1 and row-block ℓ1 norms subject to the nonnegativity constraint. Simulation results show that (i) the use of the Dual-IS and KL divergences yields better performance than the squared Euclidean distance and that (ii) the use of the Dual-IS divergence prevents from false alarms efficiently.

AB - In this paper, we present a convex-analytic approach to supervised nonnegative matrix factorization (SNMF) based on the Dual-Itakura-Saito (Dual-IS) and Kullback-Leibler (KL) divergences for music transcription. The Dual-IS and KL divergences define convex fidelity functions, whereas the IS divergence defines a nonconvex one. The SNMF problem is formulated as minimizing the divergence-based fidelity function penalized by the ℓ1 and row-block ℓ1 norms subject to the nonnegativity constraint. Simulation results show that (i) the use of the Dual-IS and KL divergences yields better performance than the squared Euclidean distance and that (ii) the use of the Dual-IS divergence prevents from false alarms efficiently.

UR - http://www.scopus.com/inward/record.url?scp=85005939755&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85005939755&partnerID=8YFLogxK

U2 - 10.1109/EUSIPCO.2016.7760426

DO - 10.1109/EUSIPCO.2016.7760426

M3 - Conference contribution

AN - SCOPUS:85005939755

VL - 2016-November

SP - 1138

EP - 1142

BT - 2016 24th European Signal Processing Conference, EUSIPCO 2016

PB - European Signal Processing Conference, EUSIPCO

T2 - 24th European Signal Processing Conference, EUSIPCO 2016

Y2 - 28 August 2016 through 2 September 2016

ER -

Supervised nonnegative matrix factorization with Dual-Itakura-Saito and Kullback-Leibler divergences for music transcription

Abstract

Other

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this