ONLINE ALGORITHMS FOR NONNEGATIVE MATRIX FACTORIZATION WITH THE ITAKURA-SAITO DIVERGENCE

被引：0

作者：

Lefevre, Augustin ^{[1
,2
]}

Bach, Francis ^{[1
]}

Fevotte, Cedric ^{[2
]}

机构：

[1] INRIA ENS Sierra Team, 23 Ave Italie, F-75013 Paris, France

[2] CNRS LTCI Telecom Paris Tech, F-75014 Paris, France

来源：

2011 IEEE WORKSHOP ON APPLICATIONS OF SIGNAL PROCESSING TO AUDIO AND ACOUSTICS (WASPAA) | 2011年

关键词：

nonnegative matrix factorization; online algorithms; Itakura-Saito divergence; blind source separation;

D O I：

暂无

中图分类号：

O42 [声学];

学科分类号：

070206 ; 082403 ;

摘要：

Nonnegative matrix factorization (NMF) is now a common tool for audio source separation. When learning NMF on large audio databases, one major drawback is that the complexity in time is O(F K N) when updating the dictionary (where (F, N) is the dimension of the input power spectrograms, and K the number of basis spectra), thus forbidding its application on signals longer than an hour. We provide an online algorithm with a complexity of O(F K) in time and memory for updates in the dictionary. We show on audio simulations that the online approach is faster for short audio signals and allows to analyze audio signals of several hours.

引用

页码：313 / 316

页数：4

共 50 条

[21] CATALOG-BASED SINGLE-CHANNEL SPEECH-MUSIC SEPARATION WITH THE ITAKURA-SAITO DIVERGENCE
Demir, Cemil
Cemgil, A. Taylan
Saraclar, Murat
[J]. 2012 PROCEEDINGS OF THE 20TH EUROPEAN SIGNAL PROCESSING CONFERENCE (EUSIPCO), 2012, : 2812 - 2816
[22] Algorithms for Nonnegative Matrix Factorization with the Kullback-Leibler Divergence
Hien, Le Thi Khanh
Gillis, Nicolas
[J]. JOURNAL OF SCIENTIFIC COMPUTING, 2021, 87 (03)
[23] ON USING THE ITAKURA-SAITO MEASURES FOR SPEECH CODER PERFORMANCE EVALUATION
JUANG, BH
[J]. AT&T BELL LABORATORIES TECHNICAL JOURNAL, 1984, 63 (08): : 1477 - 1498
[24] Sensory Evaluation of Odor Approximation Using NMF with Kullback-Leibler Divergence and Itakura-Saito Divergence in Mass Spectrum Space
Prasetyawan, Dani
Takamichi, Nakamoto
[J]. JOURNAL OF THE ELECTROCHEMICAL SOCIETY, 2020, 167 (16)
[25] Unsupervised Single-Channel Separation of Nonstationary Signals Using Gammatone Filterbank and Itakura-Saito Nonnegative Matrix Two-Dimensional Factorizations
Gao, Bin
Woo, W. L.
Dlay, S. S.
[J]. IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS I-REGULAR PAPERS, 2013, 60 (03) : 662 - 675
[26] A Study on Footstep Detection in Outdoor Environment using Itakura-Saito Distance
Tanaka, Shota
Itai, Akitoshi
Yasukawa, Hiroshi
[J]. 2014 IEEE ASIA PACIFIC CONFERENCE ON CIRCUITS AND SYSTEMS (APCCAS), 2014, : 93 - 96
[27] A contribution for the automatic sleep classification based on the Itakura-Saito spectral distance
Cardoso, Eduardo
Batista, Arnaldo
Rodrigues, Rui
Ortigueira, Manuel
Bárbara, Cristina
Martinho, Cristina
Rato, Raul
[J]. IFIP Advances in Information and Communication Technology, 2010, 314 : 374 - 381
[28] 基于Itakura-Saito散度的图像分割法
赵勇
吴成茂
[J]. 西安邮电大学学报, 2007, (01) : 80 - 83
[29] Projective Nonnegative Matrix Factorization with α-Divergence
Yang, Zhirong
Oja, Erkki
[J]. ARTIFICIAL NEURAL NETWORKS - ICANN 2009, PT I, 2009, 5768 : 20 - 29
[30] A Contribution for the Automatic Sleep Classification Based on the Itakura-Saito Spectral Distance
Cardoso, Eduardo
Batista, Arnaldo
Rodrigues, Rui
Ortigueira, Manuel
Barbara, Cristina
Martinho, Cristina
Rato, Raul
[J]. EMERGING TRENDS IN TECHNOLOGICAL INNOVATION, 2010, 314 : 374 - +

← 1 2 3 4 5 →