Please use this identifier to cite or link to this item: http://repositorio.unicamp.br/jspui/handle/REPOSIP/243282
Type: Artigo
Title: An overview of blind source separation methods for linear-quadratic and post-nonlinear mixtures
Author: Deville, Yannick
Duarte, Leonardo Tomazeli
Abstract: Whereas most blind source separation (BSS) and blind mixture identification (BMI) investigations concern linear mixtures (instantaneous or not), various recent works extended BSS and BMI to nonlinear mixing models. They especially focused on two types of models, namely linear-quadratic ones (including their bilinear and quadratic versions, and some polynomial extensions) and post-nonlinear ones. These works are particularly motivated by the associated application fields, which include remote sensing, processing of scanned images (show-through effect) and design of smart chemical and gas sensor arrays. In this paper, we provide an overview of the above two types of mixing models and of the associated BSS and/or BMI methods and applications.
Whereas most blind source separation (BSS) and blind mixture identification (BMI) investigations concern linear mixtures (instantaneous or not), various recent works extended BSS and BMI to nonlinear mixing models. They especially focused on two types of
metadata.dc.description.abstractalternative: Whereas most blind source separation (BSS) and blind mixture identification (BMI) investigations concern linear mixtures (instantaneous or not), various recent works extended BSS and BMI to nonlinear mixing models. They especially focused on two types of
Subject: Redes neurais (Ciência da computação)
Algorítmos
Country: Alemanha
Editor: Springer
Citation: An Overview Of Blind Source Separation Methods For Linear-quadratic And Post-nonlinear Mixtures. Springer-verlag Berlin, v. 9237, p. 155-167 2015.
Rights: fechado
fechado
Identifier DOI: 10.1007/978-3-319-22482-4_18
Address: https://link.springer.com/chapter/10.1007/978-3-319-22482-4_18
Date Issue: 2015
Appears in Collections:FCA - Artigos e Outros Documentos

Files in This Item:
File SizeFormat 
000363785500018.pdf257.8 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.