Academic Bibliography

 Search 200 years of publications by Ghent University researchers.
Advanced search
2 files | 2.22 MB
Add to list
  1. Search results
  2. A bio-inspired model for audio proces...

A bio-inspired model for audio processing

Tanguy Cazalets (UGent) and Joni Dambre (UGent)
(2023) 2023 RIVF International Conference on Computing and Communication Technologies (RIVF). p.458-463
Author
Organization
Project
Abstract
Homeostatic Activity Dependant Structural Plasticity (HADSP) is a recently introduced technique to generate network using structural plasticity. The algorithm use only homeostatic plasticity but let emerge principles of Hebbian learning. A previous study suggested that HADSP was able to generate networks that effectively leverage the inter-relationships between correlated time series but the idea was tested only on simple benchmarks. This paper examines HADSP's performance in speech recognition, its first application on a realistic dataset. Mimicking human hearing, a single-variable recording is transformed into a multi-variable time series through audio processing. The bio-inspired HADSP algorithm then creates a reservoir computing architecture, enhancing data representation and improving performance of the reservoir. Our principal results are that using spectral representation of the audio signal greatly improves the performance of speech recognition for echo state networks (ESNs). HADSP generated architectures show improvements in performance, corroborating the algorithm capacity to generate better reservoir connectivity.

Downloads

  • Download
    Ds754 acc.pdf
    • full text (Accepted manuscript)
    • |
    • open access
    • |
    • PDF
    • |
    • 1.08 MB
  • Download
    (...).pdf
    • full text (Published version)
    • |
    • UGent only
    • |
    • PDF
    • |
    • 1.14 MB

Citation

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

MLA
Cazalets, Tanguy, and Joni Dambre. “A Bio-Inspired Model for Audio Processing.” 2023 RIVF International Conference on Computing and Communication Technologies (RIVF), IEEE, 2023, pp. 458–63, doi:10.1109/rivf60135.2023.10471845.
APA
Cazalets, T., & Dambre, J. (2023). A bio-inspired model for audio processing. 2023 RIVF International Conference on Computing and Communication Technologies (RIVF), 458–463. https://doi.org/10.1109/rivf60135.2023.10471845
Chicago author-date
Cazalets, Tanguy, and Joni Dambre. 2023. “A Bio-Inspired Model for Audio Processing.” In 2023 RIVF International Conference on Computing and Communication Technologies (RIVF), 458–63. IEEE. https://doi.org/10.1109/rivf60135.2023.10471845.
Chicago author-date (all authors)
Cazalets, Tanguy, and Joni Dambre. 2023. “A Bio-Inspired Model for Audio Processing.” In 2023 RIVF International Conference on Computing and Communication Technologies (RIVF), 458–463. IEEE. doi:10.1109/rivf60135.2023.10471845.
Vancouver
1.
Cazalets T, Dambre J. A bio-inspired model for audio processing. In: 2023 RIVF International Conference on Computing and Communication Technologies (RIVF). IEEE; 2023. p. 458–63.
IEEE
[1]
T. Cazalets and J. Dambre, “A bio-inspired model for audio processing,” in 2023 RIVF International Conference on Computing and Communication Technologies (RIVF), Hanoi, Vietnam, 2023, pp. 458–463.
@inproceedings{01HW52R1K4CSRRF12MSRNNP4X2,
  abstract     = {{Homeostatic Activity Dependant Structural Plasticity (HADSP) is a recently introduced technique to generate network using structural plasticity. The algorithm use only homeostatic plasticity but let emerge principles of Hebbian learning. A previous study suggested that HADSP was able to generate networks that effectively leverage the inter-relationships between correlated time series but the idea was tested only on simple benchmarks. This paper examines HADSP's performance in speech recognition, its first application on a realistic dataset. Mimicking human hearing, a single-variable recording is transformed into a multi-variable time series through audio processing. The bio-inspired HADSP algorithm then creates a reservoir computing architecture, enhancing data representation and improving performance of the reservoir. Our principal results are that using spectral representation of the audio signal greatly improves the performance of speech recognition for echo state networks (ESNs). HADSP generated architectures show improvements in performance, corroborating the algorithm capacity to generate better reservoir connectivity.}},
  author       = {{Cazalets, Tanguy and Dambre, Joni}},
  booktitle    = {{2023 RIVF International Conference on Computing and Communication Technologies (RIVF)}},
  isbn         = {{9798350315844}},
  issn         = {{2473-0130}},
  language     = {{eng}},
  location     = {{Hanoi, Vietnam}},
  pages        = {{458--463}},
  publisher    = {{IEEE}},
  title        = {{A bio-inspired model for audio processing}},
  url          = {{http://doi.org/10.1109/rivf60135.2023.10471845}},
  year         = {{2023}},
}
Altmetric
View in Altmetric