Guardado en:
Detalles Bibliográficos
Autores principales: Yavuz, Mehmet Can, Yanikoglu, Berrin
Formato: Preprint
Publicado: 2023
Materias:
Acceso en línea:https://arxiv.org/abs/2312.00824
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866916238420082688
author Yavuz, Mehmet Can
Yanikoglu, Berrin
author_facet Yavuz, Mehmet Can
Yanikoglu, Berrin
contents Learning a discriminative semantic space using unlabelled and noisy data remains unaddressed in a multi-label setting. We present a contrastive self-supervised learning method which is robust to data noise, grounded in the domain of variational methods. The method (VCL) utilizes variational contrastive learning with beta-divergence to learn robustly from unlabelled datasets, including uncurated and noisy datasets. We demonstrate the effectiveness of the proposed method through rigorous experiments including linear evaluation and fine-tuning scenarios with multi-label datasets in the face understanding domain. In almost all tested scenarios, VCL surpasses the performance of state-of-the-art self-supervised methods, achieving a noteworthy increase in accuracy.
format Preprint
id arxiv_https___arxiv_org_abs_2312_00824
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Variational Self-Supervised Contrastive Learning Using Beta Divergence
Yavuz, Mehmet Can
Yanikoglu, Berrin
Computer Vision and Pattern Recognition
Machine Learning
Learning a discriminative semantic space using unlabelled and noisy data remains unaddressed in a multi-label setting. We present a contrastive self-supervised learning method which is robust to data noise, grounded in the domain of variational methods. The method (VCL) utilizes variational contrastive learning with beta-divergence to learn robustly from unlabelled datasets, including uncurated and noisy datasets. We demonstrate the effectiveness of the proposed method through rigorous experiments including linear evaluation and fine-tuning scenarios with multi-label datasets in the face understanding domain. In almost all tested scenarios, VCL surpasses the performance of state-of-the-art self-supervised methods, achieving a noteworthy increase in accuracy.
title Variational Self-Supervised Contrastive Learning Using Beta Divergence
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2312.00824