Classification of Long-Tailed Data Based on Bilateral-Branch Generative Network with Time-Supervised Strategy

In the face of the long-tailed data distribution that widely exists in real-world datasets, this paper proposes a bilateral-branch generative network model. The data of the second branch is constructed by resampling the generative network training method to improve the data quality. A bilateral-bran...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Yalin Huang, Yan-Hui Zhu, Zeng Zhigao, Yangkang Ou, Lingwei Kong
Formato:	article
Lenguaje:	EN
Publicado:	Hindawi-Wiley 2021
Materias:	Electronic computers. Computer science QA75.5-76.95
Acceso en línea:	https://doaj.org/article/ae74851cc5194039b3cbfd4bf0ff30fd
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:ae74851cc5194039b3cbfd4bf0ff30fd
record_format	dspace
spelling	oai:doaj.org-article:ae74851cc5194039b3cbfd4bf0ff30fd2021-11-15T01:19:35ZClassification of Long-Tailed Data Based on Bilateral-Branch Generative Network with Time-Supervised Strategy1099-052610.1155/2021/8667868https://doaj.org/article/ae74851cc5194039b3cbfd4bf0ff30fd2021-01-01T00:00:00Zhttp://dx.doi.org/10.1155/2021/8667868https://doaj.org/toc/1099-0526In the face of the long-tailed data distribution that widely exists in real-world datasets, this paper proposes a bilateral-branch generative network model. The data of the second branch is constructed by resampling the generative network training method to improve the data quality. A bilateral-branch network model is used to curb the risk of gradient explosion and to avoid over-fitting and under-fitting with the combined effect of different data branches. Meanwhile, Time-supervised strategy is introduced to improve the model's operational efficiency and ability to cope with extreme conditions by supervising and collaboratively controlling of the bilateral-branch generative network with time-invariant parameters. Time supervised strategy could ensure the accuracy of the model while reducing the number of iterations. Experimental results on two publicly available datasets, CIFAR10 and CIFAR100, show that the proposed method effectively improves the performance of long-tail data classification.Yalin HuangYan-Hui ZhuZeng ZhigaoYangkang OuLingwei KongHindawi-WileyarticleElectronic computers. Computer scienceQA75.5-76.95ENComplexity, Vol 2021 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Electronic computers. Computer science QA75.5-76.95
spellingShingle	Electronic computers. Computer science QA75.5-76.95 Yalin Huang Yan-Hui Zhu Zeng Zhigao Yangkang Ou Lingwei Kong Classification of Long-Tailed Data Based on Bilateral-Branch Generative Network with Time-Supervised Strategy
description	In the face of the long-tailed data distribution that widely exists in real-world datasets, this paper proposes a bilateral-branch generative network model. The data of the second branch is constructed by resampling the generative network training method to improve the data quality. A bilateral-branch network model is used to curb the risk of gradient explosion and to avoid over-fitting and under-fitting with the combined effect of different data branches. Meanwhile, Time-supervised strategy is introduced to improve the model's operational efficiency and ability to cope with extreme conditions by supervising and collaboratively controlling of the bilateral-branch generative network with time-invariant parameters. Time supervised strategy could ensure the accuracy of the model while reducing the number of iterations. Experimental results on two publicly available datasets, CIFAR10 and CIFAR100, show that the proposed method effectively improves the performance of long-tail data classification.
format	article
author	Yalin Huang Yan-Hui Zhu Zeng Zhigao Yangkang Ou Lingwei Kong
author_facet	Yalin Huang Yan-Hui Zhu Zeng Zhigao Yangkang Ou Lingwei Kong
author_sort	Yalin Huang
title	Classification of Long-Tailed Data Based on Bilateral-Branch Generative Network with Time-Supervised Strategy
title_short	Classification of Long-Tailed Data Based on Bilateral-Branch Generative Network with Time-Supervised Strategy
title_full	Classification of Long-Tailed Data Based on Bilateral-Branch Generative Network with Time-Supervised Strategy
title_fullStr	Classification of Long-Tailed Data Based on Bilateral-Branch Generative Network with Time-Supervised Strategy
title_full_unstemmed	Classification of Long-Tailed Data Based on Bilateral-Branch Generative Network with Time-Supervised Strategy
title_sort	classification of long-tailed data based on bilateral-branch generative network with time-supervised strategy
publisher	Hindawi-Wiley
publishDate	2021
url	https://doaj.org/article/ae74851cc5194039b3cbfd4bf0ff30fd
work_keys_str_mv	AT yalinhuang classificationoflongtaileddatabasedonbilateralbranchgenerativenetworkwithtimesupervisedstrategy AT yanhuizhu classificationoflongtaileddatabasedonbilateralbranchgenerativenetworkwithtimesupervisedstrategy AT zengzhigao classificationoflongtaileddatabasedonbilateralbranchgenerativenetworkwithtimesupervisedstrategy AT yangkangou classificationoflongtaileddatabasedonbilateralbranchgenerativenetworkwithtimesupervisedstrategy AT lingweikong classificationoflongtaileddatabasedonbilateralbranchgenerativenetworkwithtimesupervisedstrategy
_version_	1718428942731313152

Classification of Long-Tailed Data Based on Bilateral-Branch Generative Network with Time-Supervised Strategy

Ejemplares similares