Analysis of Scale-Variant Robust Kernel Optimization for Non-linear Least Squares Problems

In this article, we present a method for increasing adaptivity of an existing robust estimation algorithm by learning two parameters to better fit the residual distribution. The analyzed method uses these two parameters to calculate weights for Iterative Re-weighted Least Squares. This adaptive natu...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE transactions on aerospace and electronic systems 2023-06, p.1-10
Hauptverfasser: Das, Shounak, Gross, Jason N.
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page 10
container_issue
container_start_page 1
container_title IEEE transactions on aerospace and electronic systems
container_volume
creator Das, Shounak
Gross, Jason N.
description In this article, we present a method for increasing adaptivity of an existing robust estimation algorithm by learning two parameters to better fit the residual distribution. The analyzed method uses these two parameters to calculate weights for Iterative Re-weighted Least Squares. This adaptive nature of the weights can be helpful in situations where the noise level varies in the measurements. We test our algorithm first on the point cloud registration problem with synthetic data sets and LiDAR odometry with open source real-world data sets. We show that the existing approach needs an additional manual tuning of a residual scale parameter which our method directly learns from data and has similar or better performance. We further present the idea of decoupling scale and shape parameters to improve performance of the algorithm. We give detailed analysis of our algorithm along with its comparison with similar well-known algorithms from literature to show the benefits of the proposed approach.
doi_str_mv 10.1109/TAES.2023.3290142
format Article
fullrecord <record><control><sourceid>ieee_RIE</sourceid><recordid>TN_cdi_ieee_primary_10167775</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>10167775</ieee_id><sourcerecordid>10167775</sourcerecordid><originalsourceid>FETCH-ieee_primary_101677753</originalsourceid><addsrcrecordid>eNqFirsKwjAUQDMo-PwAweH-QGuSVtuOIoqgqNji4CJXuYVImtSkDvr1Org7HQ7nMDYSPBSCZ5NivsxDyWUURjLjIpYt1uVcpEEmp6LDet7fvxqncdRl57lB_fLKgy0hv6Gm4IROoWngaK9P38CGnCEN-7pRlXpjo6yB0jrYWRNoZQgdbAm_Y_54oiMPB2evmio_YO0Stafhj302Xi2LxTpQRHSpnarQvS6Ci1mSJNPoT_4ASQRBzQ</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>Analysis of Scale-Variant Robust Kernel Optimization for Non-linear Least Squares Problems</title><source>IEEE Electronic Library (IEL)</source><creator>Das, Shounak ; Gross, Jason N.</creator><creatorcontrib>Das, Shounak ; Gross, Jason N.</creatorcontrib><description>In this article, we present a method for increasing adaptivity of an existing robust estimation algorithm by learning two parameters to better fit the residual distribution. The analyzed method uses these two parameters to calculate weights for Iterative Re-weighted Least Squares. This adaptive nature of the weights can be helpful in situations where the noise level varies in the measurements. We test our algorithm first on the point cloud registration problem with synthetic data sets and LiDAR odometry with open source real-world data sets. We show that the existing approach needs an additional manual tuning of a residual scale parameter which our method directly learns from data and has similar or better performance. We further present the idea of decoupling scale and shape parameters to improve performance of the algorithm. We give detailed analysis of our algorithm along with its comparison with similar well-known algorithms from literature to show the benefits of the proposed approach.</description><identifier>ISSN: 0018-9251</identifier><identifier>DOI: 10.1109/TAES.2023.3290142</identifier><identifier>CODEN: IEARAX</identifier><language>eng</language><publisher>IEEE</publisher><subject>adaptive loss ; Cost function ; Costs ; Estimation ; Iterative methods ; iterative non-linear least squares ; Noise measurement ; Point cloud compression ; point cloud registration ; Robust estimation ; Robustness</subject><ispartof>IEEE transactions on aerospace and electronic systems, 2023-06, p.1-10</ispartof><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><orcidid>0000-0002-7771-2757 ; 0000-0002-5481-3969</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/10167775$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>314,780,784,796,27915,27916,54749</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/10167775$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Das, Shounak</creatorcontrib><creatorcontrib>Gross, Jason N.</creatorcontrib><title>Analysis of Scale-Variant Robust Kernel Optimization for Non-linear Least Squares Problems</title><title>IEEE transactions on aerospace and electronic systems</title><addtitle>T-AES</addtitle><description>In this article, we present a method for increasing adaptivity of an existing robust estimation algorithm by learning two parameters to better fit the residual distribution. The analyzed method uses these two parameters to calculate weights for Iterative Re-weighted Least Squares. This adaptive nature of the weights can be helpful in situations where the noise level varies in the measurements. We test our algorithm first on the point cloud registration problem with synthetic data sets and LiDAR odometry with open source real-world data sets. We show that the existing approach needs an additional manual tuning of a residual scale parameter which our method directly learns from data and has similar or better performance. We further present the idea of decoupling scale and shape parameters to improve performance of the algorithm. We give detailed analysis of our algorithm along with its comparison with similar well-known algorithms from literature to show the benefits of the proposed approach.</description><subject>adaptive loss</subject><subject>Cost function</subject><subject>Costs</subject><subject>Estimation</subject><subject>Iterative methods</subject><subject>iterative non-linear least squares</subject><subject>Noise measurement</subject><subject>Point cloud compression</subject><subject>point cloud registration</subject><subject>Robust estimation</subject><subject>Robustness</subject><issn>0018-9251</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><sourceid>RIE</sourceid><recordid>eNqFirsKwjAUQDMo-PwAweH-QGuSVtuOIoqgqNji4CJXuYVImtSkDvr1Org7HQ7nMDYSPBSCZ5NivsxDyWUURjLjIpYt1uVcpEEmp6LDet7fvxqncdRl57lB_fLKgy0hv6Gm4IROoWngaK9P38CGnCEN-7pRlXpjo6yB0jrYWRNoZQgdbAm_Y_54oiMPB2evmio_YO0Stafhj302Xi2LxTpQRHSpnarQvS6Ci1mSJNPoT_4ASQRBzQ</recordid><startdate>20230627</startdate><enddate>20230627</enddate><creator>Das, Shounak</creator><creator>Gross, Jason N.</creator><general>IEEE</general><scope>97E</scope><scope>RIA</scope><scope>RIE</scope><orcidid>https://orcid.org/0000-0002-7771-2757</orcidid><orcidid>https://orcid.org/0000-0002-5481-3969</orcidid></search><sort><creationdate>20230627</creationdate><title>Analysis of Scale-Variant Robust Kernel Optimization for Non-linear Least Squares Problems</title><author>Das, Shounak ; Gross, Jason N.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-ieee_primary_101677753</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>adaptive loss</topic><topic>Cost function</topic><topic>Costs</topic><topic>Estimation</topic><topic>Iterative methods</topic><topic>iterative non-linear least squares</topic><topic>Noise measurement</topic><topic>Point cloud compression</topic><topic>point cloud registration</topic><topic>Robust estimation</topic><topic>Robustness</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Das, Shounak</creatorcontrib><creatorcontrib>Gross, Jason N.</creatorcontrib><collection>IEEE All-Society Periodicals Package (ASPP) 2005-present</collection><collection>IEEE All-Society Periodicals Package (ASPP) 1998-Present</collection><collection>IEEE Electronic Library (IEL)</collection><jtitle>IEEE transactions on aerospace and electronic systems</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Das, Shounak</au><au>Gross, Jason N.</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Analysis of Scale-Variant Robust Kernel Optimization for Non-linear Least Squares Problems</atitle><jtitle>IEEE transactions on aerospace and electronic systems</jtitle><stitle>T-AES</stitle><date>2023-06-27</date><risdate>2023</risdate><spage>1</spage><epage>10</epage><pages>1-10</pages><issn>0018-9251</issn><coden>IEARAX</coden><abstract>In this article, we present a method for increasing adaptivity of an existing robust estimation algorithm by learning two parameters to better fit the residual distribution. The analyzed method uses these two parameters to calculate weights for Iterative Re-weighted Least Squares. This adaptive nature of the weights can be helpful in situations where the noise level varies in the measurements. We test our algorithm first on the point cloud registration problem with synthetic data sets and LiDAR odometry with open source real-world data sets. We show that the existing approach needs an additional manual tuning of a residual scale parameter which our method directly learns from data and has similar or better performance. We further present the idea of decoupling scale and shape parameters to improve performance of the algorithm. We give detailed analysis of our algorithm along with its comparison with similar well-known algorithms from literature to show the benefits of the proposed approach.</abstract><pub>IEEE</pub><doi>10.1109/TAES.2023.3290142</doi><orcidid>https://orcid.org/0000-0002-7771-2757</orcidid><orcidid>https://orcid.org/0000-0002-5481-3969</orcidid></addata></record>
fulltext fulltext_linktorsrc
identifier ISSN: 0018-9251
ispartof IEEE transactions on aerospace and electronic systems, 2023-06, p.1-10
issn 0018-9251
language eng
recordid cdi_ieee_primary_10167775
source IEEE Electronic Library (IEL)
subjects adaptive loss
Cost function
Costs
Estimation
Iterative methods
iterative non-linear least squares
Noise measurement
Point cloud compression
point cloud registration
Robust estimation
Robustness
title Analysis of Scale-Variant Robust Kernel Optimization for Non-linear Least Squares Problems
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-14T21%3A41%3A05IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-ieee_RIE&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Analysis%20of%20Scale-Variant%20Robust%20Kernel%20Optimization%20for%20Non-linear%20Least%20Squares%20Problems&rft.jtitle=IEEE%20transactions%20on%20aerospace%20and%20electronic%20systems&rft.au=Das,%20Shounak&rft.date=2023-06-27&rft.spage=1&rft.epage=10&rft.pages=1-10&rft.issn=0018-9251&rft.coden=IEARAX&rft_id=info:doi/10.1109/TAES.2023.3290142&rft_dat=%3Cieee_RIE%3E10167775%3C/ieee_RIE%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rft_ieee_id=10167775&rfr_iscdi=true