Deep Learning assessment of galaxy morphology in S-PLUS Data Release 1

dc.contributor.authorBom C.R.
dc.contributor.authorCortesi A.
dc.contributor.authorLucatelli G.
dc.contributor.authorDias L.O.
dc.contributor.authorSchubert P.
dc.contributor.authorOliveira Schwarz G.B.
dc.contributor.authorCardoso N.M.
dc.contributor.authorLima E.V.R.
dc.contributor.authorMendes De Oliveira C.
dc.contributor.authorSodre L.
dc.contributor.authorSmith Castelli A.V.
dc.contributor.authorFerrari F.
dc.contributor.authorDamke G.
dc.contributor.authorOverzier R.
dc.contributor.authorKanaan A.
dc.contributor.authorRibeiro T.
dc.contributor.authorSchoenell W.
dc.date.accessioned2024-03-12T19:18:57Z
dc.date.available2024-03-12T19:18:57Z
dc.date.issued2021
dc.description.abstract© 2021 The Author(s) Published by Oxford University Press on behalf of Royal Astronomical Society.The morphological diversity of galaxies is a relevant probe of galaxy evolution and cosmological structure formation, but the classification of galaxies in large sky surveys is becoming a significant challenge. We use data from the Stripe-82 area observed by the Southern Photometric Local Universe Survey (S-PLUS) in 12 optical bands, and present a catalogue of the morphologies of galaxies brighter than r = 17 mag determined both using a novel multiband morphometric fitting technique and Convolutional Neural Networks (CNNs) for computer vision. Using the CNNs, we find that, compared to our baseline results with three bands, the performance increases when using 5 broad and 3 narrow bands, but is poorer when using the full 12 band S-PLUS image set. However, the best result is still achieved with just three optical bands when using pre-trained network weights from an ImageNet data set. These results demonstrate the importance of using prior knowledge about neural network weights based on training in unrelated, extensive data sets, when available. Our catalogue contains 3274 galaxies in Stripe-82 that are not present in Galaxy Zoo 1 (GZ1), and we also provide our classifications for 4686 galaxies that were considered ambiguous in GZ1. Finally, we present a prospect of a novel way to take advantage of 12 band information for morphological classification using morphometric features, and we release a model that has been pre-trained on several bands that could be adapted for classifications using data from other surveys. The morphological catalogues are publicly available.
dc.description.firstpage1937
dc.description.issuenumber2
dc.description.lastpage1955
dc.description.volume507
dc.identifier.doi10.1093/mnras/stab1981
dc.identifier.issn1365-2966
dc.identifier.urihttps://dspace.mackenzie.br/handle/10899/34577
dc.relation.ispartofMonthly Notices of the Royal Astronomical Society
dc.rightsAcesso Restrito
dc.subject.otherlanguagegalaxies: fundamental parameters
dc.subject.otherlanguagegalaxies: structure
dc.subject.otherlanguagemethods: miscellaneous
dc.subject.otherlanguagesurveys
dc.subject.otherlanguagetechniques: image processing
dc.titleDeep Learning assessment of galaxy morphology in S-PLUS Data Release 1
dc.typeArtigo
local.scopus.citations9
local.scopus.eid2-s2.0-85116599784
local.scopus.subjectConvolutional neural network
local.scopus.subjectData set
local.scopus.subjectGalaxies: fundamental parameters
local.scopus.subjectGalaxies:structure
local.scopus.subjectMethods: miscellaneous
local.scopus.subjectMorphometrics
local.scopus.subjectNetwork weights
local.scopus.subjectOptical bands
local.scopus.subjectPhotometrics
local.scopus.subjectTechniques: image processing
local.scopus.updated2024-06-01
local.scopus.urlhttps://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85116599784&origin=inward
Arquivos