Spectogram Prediction with Neural Networks

Autores
García, Mario Alejandro; Destéfanis, Eduardo A.
Año de publicación
2018
Idioma
inglés
Tipo de recurso
documento de conferencia
Estado
versión publicada
Descripción
A neural network model for spectrogram magnitude prediction is presented. It has one convolutional layer that computes the shorttime Fourier transform. By choosing the magnitude of the spectrum as output and discarding the phase, it is possible to avoid complex number operations. The structure of the network and coefficients computation for this alternative are presented in detail. The model coefficients can be directly computed or trained with the gradient descent algorithm. In both cases, the results are satisfactory, but the obtained weights are different. An analysis of the differences is made. The main contribution of this article is to show that the proposed model is trainable. Consequently, the coefficients can be adapted to particular problems.
XIX Workshop Agentes y Sistemas Inteligentes (WASI)
Red de Universidades con Carreras en Informática (RedUNCI)
Materia
Ciencias Informáticas
discrete fourier transform
spectrogram
deep learning
convolutional neural network
Nivel de accesibilidad
acceso abierto
Condiciones de uso
http://creativecommons.org/licenses/by-nc-sa/4.0/
Repositorio
SEDICI (UNLP)
Institución
Universidad Nacional de La Plata
OAI Identificador
oai:sedici.unlp.edu.ar:10915/73033

id SEDICI_4fc6de751cf98cefa74574a41b42af88
oai_identifier_str oai:sedici.unlp.edu.ar:10915/73033
network_acronym_str SEDICI
repository_id_str 1329
network_name_str SEDICI (UNLP)
spelling Spectogram Prediction with Neural NetworksGarcía, Mario AlejandroDestéfanis, Eduardo A.Ciencias Informáticasdiscrete fourier transformspectrogramdeep learningconvolutional neural networkA neural network model for spectrogram magnitude prediction is presented. It has one convolutional layer that computes the shorttime Fourier transform. By choosing the magnitude of the spectrum as output and discarding the phase, it is possible to avoid complex number operations. The structure of the network and coefficients computation for this alternative are presented in detail. The model coefficients can be directly computed or trained with the gradient descent algorithm. In both cases, the results are satisfactory, but the obtained weights are different. An analysis of the differences is made. The main contribution of this article is to show that the proposed model is trainable. Consequently, the coefficients can be adapted to particular problems.XIX Workshop Agentes y Sistemas Inteligentes (WASI)Red de Universidades con Carreras en Informática (RedUNCI)2018-10info:eu-repo/semantics/conferenceObjectinfo:eu-repo/semantics/publishedVersionObjeto de conferenciahttp://purl.org/coar/resource_type/c_5794info:ar-repo/semantics/documentoDeConferenciaapplication/pdf42-51http://sedici.unlp.edu.ar/handle/10915/73033enginfo:eu-repo/semantics/altIdentifier/isbn/978-950-658-472-6info:eu-repo/semantics/openAccesshttp://creativecommons.org/licenses/by-nc-sa/4.0/Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0)reponame:SEDICI (UNLP)instname:Universidad Nacional de La Platainstacron:UNLP2025-09-29T11:12:13Zoai:sedici.unlp.edu.ar:10915/73033Institucionalhttp://sedici.unlp.edu.ar/Universidad públicaNo correspondehttp://sedici.unlp.edu.ar/oai/snrdalira@sedici.unlp.edu.arArgentinaNo correspondeNo correspondeNo correspondeopendoar:13292025-09-29 11:12:13.36SEDICI (UNLP) - Universidad Nacional de La Platafalse
dc.title.none.fl_str_mv Spectogram Prediction with Neural Networks
title Spectogram Prediction with Neural Networks
spellingShingle Spectogram Prediction with Neural Networks
García, Mario Alejandro
Ciencias Informáticas
discrete fourier transform
spectrogram
deep learning
convolutional neural network
title_short Spectogram Prediction with Neural Networks
title_full Spectogram Prediction with Neural Networks
title_fullStr Spectogram Prediction with Neural Networks
title_full_unstemmed Spectogram Prediction with Neural Networks
title_sort Spectogram Prediction with Neural Networks
dc.creator.none.fl_str_mv García, Mario Alejandro
Destéfanis, Eduardo A.
author García, Mario Alejandro
author_facet García, Mario Alejandro
Destéfanis, Eduardo A.
author_role author
author2 Destéfanis, Eduardo A.
author2_role author
dc.subject.none.fl_str_mv Ciencias Informáticas
discrete fourier transform
spectrogram
deep learning
convolutional neural network
topic Ciencias Informáticas
discrete fourier transform
spectrogram
deep learning
convolutional neural network
dc.description.none.fl_txt_mv A neural network model for spectrogram magnitude prediction is presented. It has one convolutional layer that computes the shorttime Fourier transform. By choosing the magnitude of the spectrum as output and discarding the phase, it is possible to avoid complex number operations. The structure of the network and coefficients computation for this alternative are presented in detail. The model coefficients can be directly computed or trained with the gradient descent algorithm. In both cases, the results are satisfactory, but the obtained weights are different. An analysis of the differences is made. The main contribution of this article is to show that the proposed model is trainable. Consequently, the coefficients can be adapted to particular problems.
XIX Workshop Agentes y Sistemas Inteligentes (WASI)
Red de Universidades con Carreras en Informática (RedUNCI)
description A neural network model for spectrogram magnitude prediction is presented. It has one convolutional layer that computes the shorttime Fourier transform. By choosing the magnitude of the spectrum as output and discarding the phase, it is possible to avoid complex number operations. The structure of the network and coefficients computation for this alternative are presented in detail. The model coefficients can be directly computed or trained with the gradient descent algorithm. In both cases, the results are satisfactory, but the obtained weights are different. An analysis of the differences is made. The main contribution of this article is to show that the proposed model is trainable. Consequently, the coefficients can be adapted to particular problems.
publishDate 2018
dc.date.none.fl_str_mv 2018-10
dc.type.none.fl_str_mv info:eu-repo/semantics/conferenceObject
info:eu-repo/semantics/publishedVersion
Objeto de conferencia
http://purl.org/coar/resource_type/c_5794
info:ar-repo/semantics/documentoDeConferencia
format conferenceObject
status_str publishedVersion
dc.identifier.none.fl_str_mv http://sedici.unlp.edu.ar/handle/10915/73033
url http://sedici.unlp.edu.ar/handle/10915/73033
dc.language.none.fl_str_mv eng
language eng
dc.relation.none.fl_str_mv info:eu-repo/semantics/altIdentifier/isbn/978-950-658-472-6
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
http://creativecommons.org/licenses/by-nc-sa/4.0/
Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0)
eu_rights_str_mv openAccess
rights_invalid_str_mv http://creativecommons.org/licenses/by-nc-sa/4.0/
Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0)
dc.format.none.fl_str_mv application/pdf
42-51
dc.source.none.fl_str_mv reponame:SEDICI (UNLP)
instname:Universidad Nacional de La Plata
instacron:UNLP
reponame_str SEDICI (UNLP)
collection SEDICI (UNLP)
instname_str Universidad Nacional de La Plata
instacron_str UNLP
institution UNLP
repository.name.fl_str_mv SEDICI (UNLP) - Universidad Nacional de La Plata
repository.mail.fl_str_mv alira@sedici.unlp.edu.ar
_version_ 1844615992502648832
score 13.070432