Academic Bibliography

 Search 200 years of publications by Ghent University researchers.
Advanced search
1 file | 863.29 KB
Add to list
  1. Search results
  2. Sigmoidal NMFD : convolutional NMF wi...

Sigmoidal NMFD : convolutional NMF with saturating activations for drum mixture decomposition

Len Vande Veire, Cedric De Boom and Tijl De Bie (UGent)
(2021) ELECTRONICS. 10(3).
Author
Organization
Project
Abstract
In many types of music, percussion plays an essential role to establish the rhythm and the groove of the music. Algorithms that can decompose the percussive signal into its constituent components would therefore be very useful, as they would enable many analytical and creative applications. This paper describes a method for the unsupervised decomposition of percussive recordings, building on the non-negative matrix factor deconvolution (NMFD) algorithm. Given a percussive music recording, NMFD discovers a dictionary of time-varying spectral templates and corresponding activation functions, representing its constituent sounds and their positions in the mix. We observe, however, that the activation functions discovered using NMFD do not show the expected impulse-like behavior for percussive instruments. We therefore enforce this behavior by specifying that the activations should take on binary values: either an instrument is hit, or it is not. To this end, we rewrite the activations as the output of a sigmoidal function, multiplied with a per-component amplitude factor. We furthermore define a regularization term that biases the decomposition to solutions with saturated activations, leading to the desired binary behavior. We evaluate several optimization strategies and techniques that are designed to avoid poor local minima. We show that incentivizing the activations to be binary indeed leads to the desired impulse-like behavior, and that the resulting components are better separated, leading to more interpretable decompositions.
Keywords
NMFD, automatic drum transcription, automatic drum mixture, decomposition, regularization

Downloads

  • Download
    7845.pdf
    • full text (Published version)
    • |
    • open access
    • |
    • PDF
    • |
    • 863.29 KB

Citation

Please use this url to cite or link to this publication:

MLA
Vande Veire, Len, et al. “Sigmoidal NMFD : Convolutional NMF with Saturating Activations for Drum Mixture Decomposition.” ELECTRONICS, vol. 10, no. 3, 2021, doi:10.3390/electronics10030284.
APA
Vande Veire, L., De Boom, C., & De Bie, T. (2021). Sigmoidal NMFD : convolutional NMF with saturating activations for drum mixture decomposition. ELECTRONICS, 10(3). https://doi.org/10.3390/electronics10030284
Chicago author-date
Vande Veire, Len, Cedric De Boom, and Tijl De Bie. 2021. “Sigmoidal NMFD : Convolutional NMF with Saturating Activations for Drum Mixture Decomposition.” ELECTRONICS 10 (3). https://doi.org/10.3390/electronics10030284.
Chicago author-date (all authors)
Vande Veire, Len, Cedric De Boom, and Tijl De Bie. 2021. “Sigmoidal NMFD : Convolutional NMF with Saturating Activations for Drum Mixture Decomposition.” ELECTRONICS 10 (3). doi:10.3390/electronics10030284.
Vancouver
1.
Vande Veire L, De Boom C, De Bie T. Sigmoidal NMFD : convolutional NMF with saturating activations for drum mixture decomposition. ELECTRONICS. 2021;10(3).
IEEE
[1]
L. Vande Veire, C. De Boom, and T. De Bie, “Sigmoidal NMFD : convolutional NMF with saturating activations for drum mixture decomposition,” ELECTRONICS, vol. 10, no. 3, 2021.
@article{8697656,
  abstract     = {{In many types of music, percussion plays an essential role to establish the rhythm and the groove of the music. Algorithms that can decompose the percussive signal into its constituent components would therefore be very useful, as they would enable many analytical and creative applications. This paper describes a method for the unsupervised decomposition of percussive recordings, building on the non-negative matrix factor deconvolution (NMFD) algorithm. Given a percussive music recording, NMFD discovers a dictionary of time-varying spectral templates and corresponding activation functions, representing its constituent sounds and their positions in the mix. We observe, however, that the activation functions discovered using NMFD do not show the expected impulse-like behavior for percussive instruments. We therefore enforce this behavior by specifying that the activations should take on binary values: either an instrument is hit, or it is not. To this end, we rewrite the activations as the output of a sigmoidal function, multiplied with a per-component amplitude factor. We furthermore define a regularization term that biases the decomposition to solutions with saturated activations, leading to the desired binary behavior. We evaluate several optimization strategies and techniques that are designed to avoid poor local minima. We show that incentivizing the activations to be binary indeed leads to the desired impulse-like behavior, and that the resulting components are better separated, leading to more interpretable decompositions.}},
  articleno    = {{284}},
  author       = {{Vande Veire, Len and De Boom, Cedric and De Bie, Tijl}},
  issn         = {{2079-9292}},
  journal      = {{ELECTRONICS}},
  keywords     = {{NMFD,automatic drum transcription,automatic drum mixture,decomposition,regularization}},
  language     = {{eng}},
  number       = {{3}},
  pages        = {{26}},
  title        = {{Sigmoidal NMFD : convolutional NMF with saturating activations for drum mixture decomposition}},
  url          = {{http://doi.org/10.3390/electronics10030284}},
  volume       = {{10}},
  year         = {{2021}},
}
Altmetric
View in Altmetric
Web of Science
Times cited: