Feedback Module Based Convolution Neural Networks for Sound Event Classification

Sound event classification is starting to receive a lot of attention over the recent years in the field of audio processing because of open datasets, which are recorded in various conditions, and the introduction of challenges. To use the sound event classification model in the wild, it is needed to...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Gwantae Kim, David K. Han, Hanseok Ko
Formato:	article
Lenguaje:	EN
Publicado:	IEEE 2021
Materias:	Dual-path residual network feedback module recurrent inference sound event classification Electrical engineering. Electronics. Nuclear engineering TK1-9971
Acceso en línea:	https://doaj.org/article/ea48d9131f1349c789ad9935372f1aaa
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:ea48d9131f1349c789ad9935372f1aaa
record_format	dspace
spelling	oai:doaj.org-article:ea48d9131f1349c789ad9935372f1aaa2021-11-18T00:06:50ZFeedback Module Based Convolution Neural Networks for Sound Event Classification2169-353610.1109/ACCESS.2021.3126004https://doaj.org/article/ea48d9131f1349c789ad9935372f1aaa2021-01-01T00:00:00Zhttps://ieeexplore.ieee.org/document/9605665/https://doaj.org/toc/2169-3536Sound event classification is starting to receive a lot of attention over the recent years in the field of audio processing because of open datasets, which are recorded in various conditions, and the introduction of challenges. To use the sound event classification model in the wild, it is needed to be independent of recording conditions. Therefore, a more generalized model, that can be trained and tested in various recording conditions, must be researched. This paper presents a deep neural network with a dual-path frequency residual network and feedback modules for sound event classification. Most deep neural network based approaches for sound event classification use feed-forward models and train with a single classification result. Although these methods are simple to implement and deliver reasonable results, the integration of recurrent inference based methods has shown potential for classification and generalization performance improvements. We propose a weighted recurrent inference based model by employing cascading feedback modules for sound event classification. In our experiments, it is shown that the proposed method outperforms traditional approaches in indoor and outdoor conditions by 1.94% and 3.26%, respectively.Gwantae KimDavid K. HanHanseok KoIEEEarticleDual-path residual networkfeedback modulerecurrent inferencesound event classificationElectrical engineering. Electronics. Nuclear engineeringTK1-9971ENIEEE Access, Vol 9, Pp 150993-151003 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Dual-path residual network feedback module recurrent inference sound event classification Electrical engineering. Electronics. Nuclear engineering TK1-9971
spellingShingle	Dual-path residual network feedback module recurrent inference sound event classification Electrical engineering. Electronics. Nuclear engineering TK1-9971 Gwantae Kim David K. Han Hanseok Ko Feedback Module Based Convolution Neural Networks for Sound Event Classification
description	Sound event classification is starting to receive a lot of attention over the recent years in the field of audio processing because of open datasets, which are recorded in various conditions, and the introduction of challenges. To use the sound event classification model in the wild, it is needed to be independent of recording conditions. Therefore, a more generalized model, that can be trained and tested in various recording conditions, must be researched. This paper presents a deep neural network with a dual-path frequency residual network and feedback modules for sound event classification. Most deep neural network based approaches for sound event classification use feed-forward models and train with a single classification result. Although these methods are simple to implement and deliver reasonable results, the integration of recurrent inference based methods has shown potential for classification and generalization performance improvements. We propose a weighted recurrent inference based model by employing cascading feedback modules for sound event classification. In our experiments, it is shown that the proposed method outperforms traditional approaches in indoor and outdoor conditions by 1.94% and 3.26%, respectively.
format	article
author	Gwantae Kim David K. Han Hanseok Ko
author_facet	Gwantae Kim David K. Han Hanseok Ko
author_sort	Gwantae Kim
title	Feedback Module Based Convolution Neural Networks for Sound Event Classification
title_short	Feedback Module Based Convolution Neural Networks for Sound Event Classification
title_full	Feedback Module Based Convolution Neural Networks for Sound Event Classification
title_fullStr	Feedback Module Based Convolution Neural Networks for Sound Event Classification
title_full_unstemmed	Feedback Module Based Convolution Neural Networks for Sound Event Classification
title_sort	feedback module based convolution neural networks for sound event classification
publisher	IEEE
publishDate	2021
url	https://doaj.org/article/ea48d9131f1349c789ad9935372f1aaa
work_keys_str_mv	AT gwantaekim feedbackmodulebasedconvolutionneuralnetworksforsoundeventclassification AT davidkhan feedbackmodulebasedconvolutionneuralnetworksforsoundeventclassification AT hanseokko feedbackmodulebasedconvolutionneuralnetworksforsoundeventclassification
_version_	1718425258233430016

Feedback Module Based Convolution Neural Networks for Sound Event Classification

Ejemplares similares