TWO-WAY METRIC LEARNING WITH MAJORITY AND MINORITY SUBSETS FOR CLASSIFICATION OF LARGE EXTREMELY IMBALANCED FACE DATASET

This paper proposes a new learning methodology involving deep features and two-way metric learning for large, extremely imbalanced face datasets where the number of minority classes and the imbalance ratio are both very high. The problem arises because the faces of some celebrities, being more popul...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Ashu Kaushik, Seba Susan
Formato:	article
Lenguaje:	EN
Publicado:	Scientific Research Support Fund of Jordan (SRSF) and Princess Sumaya University for Technology (PSUT) 2021
Materias:	face recognition metric learning vgg-face deep learning imbalanced learning extremely imbalanced dataset Information technology T58.5-58.64 Electronic computers. Computer science QA75.5-76.95
Acceso en línea:	https://doaj.org/article/7e26b2cef55244e5a6807707f93b2e55
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:7e26b2cef55244e5a6807707f93b2e55
record_format	dspace
spelling	oai:doaj.org-article:7e26b2cef55244e5a6807707f93b2e552021-12-03T07:32:06ZTWO-WAY METRIC LEARNING WITH MAJORITY AND MINORITY SUBSETS FOR CLASSIFICATION OF LARGE EXTREMELY IMBALANCED FACE DATASET2413-935110.5455/jjcit.71-1626417940https://doaj.org/article/7e26b2cef55244e5a6807707f93b2e552021-12-01T00:00:00Zhttp://www.ejmanager.com/fulltextpdf.php?mno=98032https://doaj.org/toc/2413-9351This paper proposes a new learning methodology involving deep features and two-way metric learning for large, extremely imbalanced face datasets where the number of minority classes and the imbalance ratio are both very high. The problem arises because the faces of some celebrities, being more popular, are readily available in social media and the internet, while the faces of some relatively lesser-known personalities are fewer in number. Resampling being impractical in this scenario, we propose metric learning as the tool for mitigating the class-imbalance problem prior to the classification stage. To reduce the computational overhead associated with metric learning, we separately conduct weakly supervised metric learning with majority and minority class subsets, a process that we call two-way metric learning. Transformation matrices learnt from the majority and minority subsets are used to transform the entire input space twice. The test sample in the transformed space is assigned the class of its nearest neighbor in the training set of the twice-transformed input space. Deep features derived from the state-of-the-art pre-trained deep network VGG-Face form the input space, and the aggregate cosine similarity measure is used to find the closest neighbor in the training set of the twice-transformed input space. Experiments on the benchmark LFW face database having 1680 classes of celebrity faces prove that the proposed methodology is more effective than existing methods for the classification of large, extremely imbalanced face datasets. The classification accuracies of the minority classes are especially found to be boosted which is a rare accomplishment among existing methods for imbalanced learning in deep frameworks. [JJCIT 2021; 7(4.000): 337-348]Ashu KaushikSeba SusanScientific Research Support Fund of Jordan (SRSF) and Princess Sumaya University for Technology (PSUT)articleface recognitionmetric learningvgg-facedeep learningimbalanced learningextremely imbalanced datasetInformation technologyT58.5-58.64Electronic computers. Computer scienceQA75.5-76.95ENJordanian Journal of Computers and Information Technology , Vol 7, Iss 4, Pp 337-348 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	face recognition metric learning vgg-face deep learning imbalanced learning extremely imbalanced dataset Information technology T58.5-58.64 Electronic computers. Computer science QA75.5-76.95
spellingShingle	face recognition metric learning vgg-face deep learning imbalanced learning extremely imbalanced dataset Information technology T58.5-58.64 Electronic computers. Computer science QA75.5-76.95 Ashu Kaushik Seba Susan TWO-WAY METRIC LEARNING WITH MAJORITY AND MINORITY SUBSETS FOR CLASSIFICATION OF LARGE EXTREMELY IMBALANCED FACE DATASET
description	This paper proposes a new learning methodology involving deep features and two-way metric learning for large, extremely imbalanced face datasets where the number of minority classes and the imbalance ratio are both very high. The problem arises because the faces of some celebrities, being more popular, are readily available in social media and the internet, while the faces of some relatively lesser-known personalities are fewer in number. Resampling being impractical in this scenario, we propose metric learning as the tool for mitigating the class-imbalance problem prior to the classification stage. To reduce the computational overhead associated with metric learning, we separately conduct weakly supervised metric learning with majority and minority class subsets, a process that we call two-way metric learning. Transformation matrices learnt from the majority and minority subsets are used to transform the entire input space twice. The test sample in the transformed space is assigned the class of its nearest neighbor in the training set of the twice-transformed input space. Deep features derived from the state-of-the-art pre-trained deep network VGG-Face form the input space, and the aggregate cosine similarity measure is used to find the closest neighbor in the training set of the twice-transformed input space. Experiments on the benchmark LFW face database having 1680 classes of celebrity faces prove that the proposed methodology is more effective than existing methods for the classification of large, extremely imbalanced face datasets. The classification accuracies of the minority classes are especially found to be boosted which is a rare accomplishment among existing methods for imbalanced learning in deep frameworks. [JJCIT 2021; 7(4.000): 337-348]
format	article
author	Ashu Kaushik Seba Susan
author_facet	Ashu Kaushik Seba Susan
author_sort	Ashu Kaushik
title	TWO-WAY METRIC LEARNING WITH MAJORITY AND MINORITY SUBSETS FOR CLASSIFICATION OF LARGE EXTREMELY IMBALANCED FACE DATASET
title_short	TWO-WAY METRIC LEARNING WITH MAJORITY AND MINORITY SUBSETS FOR CLASSIFICATION OF LARGE EXTREMELY IMBALANCED FACE DATASET
title_full	TWO-WAY METRIC LEARNING WITH MAJORITY AND MINORITY SUBSETS FOR CLASSIFICATION OF LARGE EXTREMELY IMBALANCED FACE DATASET
title_fullStr	TWO-WAY METRIC LEARNING WITH MAJORITY AND MINORITY SUBSETS FOR CLASSIFICATION OF LARGE EXTREMELY IMBALANCED FACE DATASET
title_full_unstemmed	TWO-WAY METRIC LEARNING WITH MAJORITY AND MINORITY SUBSETS FOR CLASSIFICATION OF LARGE EXTREMELY IMBALANCED FACE DATASET
title_sort	two-way metric learning with majority and minority subsets for classification of large extremely imbalanced face dataset
publisher	Scientific Research Support Fund of Jordan (SRSF) and Princess Sumaya University for Technology (PSUT)
publishDate	2021
url	https://doaj.org/article/7e26b2cef55244e5a6807707f93b2e55
work_keys_str_mv	AT ashukaushik twowaymetriclearningwithmajorityandminoritysubsetsforclassificationoflargeextremelyimbalancedfacedataset AT sebasusan twowaymetriclearningwithmajorityandminoritysubsetsforclassificationoflargeextremelyimbalancedfacedataset
_version_	1718373820068265984

TWO-WAY METRIC LEARNING WITH MAJORITY AND MINORITY SUBSETS FOR CLASSIFICATION OF LARGE EXTREMELY IMBALANCED FACE DATASET

Ejemplares similares