Denoising Medium Resolution Stellar Spectra With Neural Networks

IF 1.1 4区 物理与天体物理 Q3 ASTRONOMY & ASTROPHYSICS
Balázs Pál, László Dobos
{"title":"Denoising Medium Resolution Stellar Spectra With Neural Networks","authors":"Balázs Pál,&nbsp;László Dobos","doi":"10.1002/asna.20240049","DOIUrl":null,"url":null,"abstract":"<div>\n \n <p>We trained denoiser autoencoding neural networks on medium resolution simulated optical spectra of late-type stars to demonstrate that the reconstruction of the original flux is possible at a typical relative error of a fraction of a percent down to a typical signal-to-noise ratio of <span></span><math>\n <semantics>\n <mrow>\n <mn>10</mn>\n </mrow>\n <annotation>$$ 10 $$</annotation>\n </semantics></math> per pixel. We show that relatively simple networks are capable of learning the characteristics of stellar spectra while still flexible enough to adapt to different values of extinction and fluxing imperfections that modifies the overall shape of the continuum, as well as to different values of Doppler shift. Denoised spectra can be used to find initial values for traditional stellar template fitting algorithms and—since evaluation of pretrained neural networks is significantly faster than traditional template fitting—denoiser networks can be useful when a fast analysis of the noisy spectrum is necessary, for example during observations, between individual exposures.</p>\n </div>","PeriodicalId":55442,"journal":{"name":"Astronomische Nachrichten","volume":"345 9-10","pages":""},"PeriodicalIF":1.1000,"publicationDate":"2024-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Astronomische Nachrichten","FirstCategoryId":"101","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1002/asna.20240049","RegionNum":4,"RegionCategory":"物理与天体物理","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"ASTRONOMY & ASTROPHYSICS","Score":null,"Total":0}
引用次数: 0

Abstract

We trained denoiser autoencoding neural networks on medium resolution simulated optical spectra of late-type stars to demonstrate that the reconstruction of the original flux is possible at a typical relative error of a fraction of a percent down to a typical signal-to-noise ratio of 10 $$ 10 $$ per pixel. We show that relatively simple networks are capable of learning the characteristics of stellar spectra while still flexible enough to adapt to different values of extinction and fluxing imperfections that modifies the overall shape of the continuum, as well as to different values of Doppler shift. Denoised spectra can be used to find initial values for traditional stellar template fitting algorithms and—since evaluation of pretrained neural networks is significantly faster than traditional template fitting—denoiser networks can be useful when a fast analysis of the noisy spectrum is necessary, for example during observations, between individual exposures.

用神经网络去噪中分辨率恒星光谱
我们在中分辨率模拟晚型恒星的光谱上训练去噪自编码神经网络,以证明原始通量的重建是可能的,典型的相对误差为百分之零点几,典型的信噪比为每像素10 $$ 10 $$。我们表明,相对简单的网络能够学习恒星光谱的特征,同时仍然足够灵活,以适应不同的消光值和通量缺陷,这些缺陷改变了连续体的整体形状,以及不同的多普勒频移值。去噪光谱可用于寻找传统恒星模板拟合算法的初始值,并且由于预训练神经网络的评估比传统模板拟合要快得多,因此当需要对噪声光谱进行快速分析时,例如在观测期间,在个别曝光之间,去噪网络是有用的。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
Astronomische Nachrichten
Astronomische Nachrichten 地学天文-天文与天体物理
CiteScore
1.80
自引率
11.10%
发文量
57
审稿时长
4-8 weeks
期刊介绍: Astronomische Nachrichten, founded in 1821 by H. C. Schumacher, is the oldest astronomical journal worldwide still being published. Famous astronomical discoveries and important papers on astronomy and astrophysics published in more than 300 volumes of the journal give an outstanding representation of the progress of astronomical research over the last 180 years. Today, Astronomical Notes/ Astronomische Nachrichten publishes articles in the field of observational and theoretical astrophysics and related topics in solar-system and solar physics. Additional, papers on astronomical instrumentation ground-based and space-based as well as papers about numerical astrophysical techniques and supercomputer modelling are covered. Papers can be completed by short video sequences in the electronic version. Astronomical Notes/ Astronomische Nachrichten also publishes special issues of meeting proceedings.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信