Optimizing Few-Shot Learning Based on Variational Autoencoders

Despite the importance of few-shot learning, the lack of labeled training data in the real world makes it extremely challenging for existing machine learning methods because this limited dataset does not well represent the data variance. In this research, we suggest employing a generative approach u...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Ruoqi Wei, Ausif Mahmood
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	deep learning variational autoencoders data representation learning generative models unsupervised learning few shot learning Science Q Astrophysics QB460-466 Physics QC1-999
Acceso en línea:	https://doaj.org/article/df412ec1c61f467f80127d304633eb2e
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:df412ec1c61f467f80127d304633eb2e
record_format	dspace
spelling	oai:doaj.org-article:df412ec1c61f467f80127d304633eb2e2021-11-25T17:29:14ZOptimizing Few-Shot Learning Based on Variational Autoencoders10.3390/e231113901099-4300https://doaj.org/article/df412ec1c61f467f80127d304633eb2e2021-10-01T00:00:00Zhttps://www.mdpi.com/1099-4300/23/11/1390https://doaj.org/toc/1099-4300Despite the importance of few-shot learning, the lack of labeled training data in the real world makes it extremely challenging for existing machine learning methods because this limited dataset does not well represent the data variance. In this research, we suggest employing a generative approach using variational autoencoders (VAEs), which can be used specifically to optimize few-shot learning tasks by generating new samples with more intra-class variations on the Labeled Faces in the Wild (LFW) dataset. The purpose of our research is to increase the size of the training dataset using various methods to improve the accuracy and robustness of the few-shot face recognition. Specifically, we employ the VAE generator to increase the size of the training dataset, including the basic and the novel sets while utilizing transfer learning as the backend. Based on extensive experimental research, we analyze various data augmentation methods to observe how each method affects the accuracy of face recognition. The face generation method based on VAEs with perceptual loss can effectively improve the recognition accuracy rate to 96.47% using both the base and the novel sets.Ruoqi WeiAusif MahmoodMDPI AGarticledeep learningvariational autoencodersdata representation learninggenerative modelsunsupervised learningfew shot learningScienceQAstrophysicsQB460-466PhysicsQC1-999ENEntropy, Vol 23, Iss 1390, p 1390 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	deep learning variational autoencoders data representation learning generative models unsupervised learning few shot learning Science Q Astrophysics QB460-466 Physics QC1-999
spellingShingle	deep learning variational autoencoders data representation learning generative models unsupervised learning few shot learning Science Q Astrophysics QB460-466 Physics QC1-999 Ruoqi Wei Ausif Mahmood Optimizing Few-Shot Learning Based on Variational Autoencoders
description	Despite the importance of few-shot learning, the lack of labeled training data in the real world makes it extremely challenging for existing machine learning methods because this limited dataset does not well represent the data variance. In this research, we suggest employing a generative approach using variational autoencoders (VAEs), which can be used specifically to optimize few-shot learning tasks by generating new samples with more intra-class variations on the Labeled Faces in the Wild (LFW) dataset. The purpose of our research is to increase the size of the training dataset using various methods to improve the accuracy and robustness of the few-shot face recognition. Specifically, we employ the VAE generator to increase the size of the training dataset, including the basic and the novel sets while utilizing transfer learning as the backend. Based on extensive experimental research, we analyze various data augmentation methods to observe how each method affects the accuracy of face recognition. The face generation method based on VAEs with perceptual loss can effectively improve the recognition accuracy rate to 96.47% using both the base and the novel sets.
format	article
author	Ruoqi Wei Ausif Mahmood
author_facet	Ruoqi Wei Ausif Mahmood
author_sort	Ruoqi Wei
title	Optimizing Few-Shot Learning Based on Variational Autoencoders
title_short	Optimizing Few-Shot Learning Based on Variational Autoencoders
title_full	Optimizing Few-Shot Learning Based on Variational Autoencoders
title_fullStr	Optimizing Few-Shot Learning Based on Variational Autoencoders
title_full_unstemmed	Optimizing Few-Shot Learning Based on Variational Autoencoders
title_sort	optimizing few-shot learning based on variational autoencoders
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/df412ec1c61f467f80127d304633eb2e
work_keys_str_mv	AT ruoqiwei optimizingfewshotlearningbasedonvariationalautoencoders AT ausifmahmood optimizingfewshotlearningbasedonvariationalautoencoders
_version_	1718412317417275392

Optimizing Few-Shot Learning Based on Variational Autoencoders

Ejemplares similares