Spline Error Weighting for Robust Visual-Inertial Fusion

In this paper we derive and test a probability-based weighting that can balance residuals of different types in spline fitting. In contrast to previous formulations, the proposed spline error weighting scheme also incorporates a prediction of the approximation error of the spline fit. We demonstrate...

Full description

Saved in:
Bibliographic Details
Main Authors: Ovren, Hannes, Forssen, Per-Erik
Format: Conference Proceeding
Language:English
Subjects:
Approximation error
Cameras
Discrete Fourier transforms
Splines (mathematics)
Trajectory
Videos
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by
cites
container_end_page 329
container_issue
container_start_page 321
container_title
container_volume
creator Ovren, Hannes
Forssen, Per-Erik
description In this paper we derive and test a probability-based weighting that can balance residuals of different types in spline fitting. In contrast to previous formulations, the proposed spline error weighting scheme also incorporates a prediction of the approximation error of the spline fit. We demonstrate the effectiveness of the prediction in a synthetic experiment, and apply it to visual-inertial fusion on rolling shutter cameras. This results in a method that can estimate 3D structure with metric scale on generic first-person videos. We also propose a quality measure for spline fitting, that can be used to automatically select the knot spacing. Experiments verify that the obtained trajectory quality corresponds well with the requested quality. Finally, by linearly scaling the weights, we show that the proposed spline error weighting minimizes the estimation errors on real sequences, in terms of scale and end-point errors.
doi_str_mv 10.1109/CVPR.2018.00041
format conference_proceeding
fullrecord <record><control><sourceid>swepub_CHZPO</sourceid><recordid>TN_cdi_ieee_primary_8578139</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>8578139</ieee_id><sourcerecordid>oai_DiVA_org_liu_149495</sourcerecordid><originalsourceid>FETCH-LOGICAL-i213t-c1f3f54fa8c4181cbdfa073971f8e493014efa7c4f3e7be5ff3b2e4bdcbc9f863</originalsourceid><addsrcrecordid>eNpVkMFLwzAYxaMiOObOHrz0H2jN16Rpchx1m4OBMnUeS9J9qZHalqRF_O8tTARPj8f7vXd4hNwATQCouisOT_skpSATSimHM7JQuYSMSSF4StU5mQEVLBYK1MW_DMQVWYTwMdVSIZnk2YzI575xLUYr7zsfvaGr3wfX1pGd3L4zYxiigwujbuJti35wuonWY3Bde00urW4CLn51Tl7Xq5fiId49brbFche7FNgQV2CZzbjVsuIgoTJHq2nOVA5WIleMAker84pbhrnBzFpmUuTmWJlKWSnYnMSn3fCF_WjK3rtP7b_LTrvy3h2WZefrsnFjCVxxlU387Yl3iPhHy2y6gSn2A1XOWuM</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>Spline Error Weighting for Robust Visual-Inertial Fusion</title><source>IEEE Xplore All Conference Series</source><creator>Ovren, Hannes ; Forssen, Per-Erik</creator><creatorcontrib>Ovren, Hannes ; Forssen, Per-Erik</creatorcontrib><description>In this paper we derive and test a probability-based weighting that can balance residuals of different types in spline fitting. In contrast to previous formulations, the proposed spline error weighting scheme also incorporates a prediction of the approximation error of the spline fit. We demonstrate the effectiveness of the prediction in a synthetic experiment, and apply it to visual-inertial fusion on rolling shutter cameras. This results in a method that can estimate 3D structure with metric scale on generic first-person videos. We also propose a quality measure for spline fitting, that can be used to automatically select the knot spacing. Experiments verify that the obtained trajectory quality corresponds well with the requested quality. Finally, by linearly scaling the weights, we show that the proposed spline error weighting minimizes the estimation errors on real sequences, in terms of scale and end-point errors.</description><identifier>ISBN: 9781538664216</identifier><identifier>ISBN: 1538664216</identifier><identifier>ISBN: 9781538664209</identifier><identifier>ISBN: 1538664208</identifier><identifier>EISSN: 1063-6919</identifier><identifier>EISBN: 9781538664209</identifier><identifier>EISBN: 1538664208</identifier><identifier>DOI: 10.1109/CVPR.2018.00041</identifier><identifier>CODEN: IEEPAD</identifier><language>eng</language><publisher>IEEE</publisher><subject>Approximation error ; Cameras ; Discrete Fourier transforms ; Splines (mathematics) ; Trajectory ; Videos</subject><ispartof>2018 IEEE CVF Conference on Computer Vision and Pattern Recognition, 2018, p.321-329</ispartof><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/8578139$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,776,780,785,786,881,4036,4037,27902,54530,54907</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/8578139$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc><backlink>$$Uhttps://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-149495$$DView record from Swedish Publication Index$$Hfree_for_read</backlink></links><search><creatorcontrib>Ovren, Hannes</creatorcontrib><creatorcontrib>Forssen, Per-Erik</creatorcontrib><title>Spline Error Weighting for Robust Visual-Inertial Fusion</title><title>2018 IEEE CVF Conference on Computer Vision and Pattern Recognition</title><addtitle>CVPR</addtitle><description>In this paper we derive and test a probability-based weighting that can balance residuals of different types in spline fitting. In contrast to previous formulations, the proposed spline error weighting scheme also incorporates a prediction of the approximation error of the spline fit. We demonstrate the effectiveness of the prediction in a synthetic experiment, and apply it to visual-inertial fusion on rolling shutter cameras. This results in a method that can estimate 3D structure with metric scale on generic first-person videos. We also propose a quality measure for spline fitting, that can be used to automatically select the knot spacing. Experiments verify that the obtained trajectory quality corresponds well with the requested quality. Finally, by linearly scaling the weights, we show that the proposed spline error weighting minimizes the estimation errors on real sequences, in terms of scale and end-point errors.</description><subject>Approximation error</subject><subject>Cameras</subject><subject>Discrete Fourier transforms</subject><subject>Splines (mathematics)</subject><subject>Trajectory</subject><subject>Videos</subject><issn>1063-6919</issn><isbn>9781538664216</isbn><isbn>1538664216</isbn><isbn>9781538664209</isbn><isbn>1538664208</isbn><isbn>9781538664209</isbn><isbn>1538664208</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>2018</creationdate><recordtype>conference_proceeding</recordtype><sourceid>6IE</sourceid><recordid>eNpVkMFLwzAYxaMiOObOHrz0H2jN16Rpchx1m4OBMnUeS9J9qZHalqRF_O8tTARPj8f7vXd4hNwATQCouisOT_skpSATSimHM7JQuYSMSSF4StU5mQEVLBYK1MW_DMQVWYTwMdVSIZnk2YzI575xLUYr7zsfvaGr3wfX1pGd3L4zYxiigwujbuJti35wuonWY3Bde00urW4CLn51Tl7Xq5fiId49brbFche7FNgQV2CZzbjVsuIgoTJHq2nOVA5WIleMAker84pbhrnBzFpmUuTmWJlKWSnYnMSn3fCF_WjK3rtP7b_LTrvy3h2WZefrsnFjCVxxlU387Yl3iPhHy2y6gSn2A1XOWuM</recordid><startdate>201806</startdate><enddate>201806</enddate><creator>Ovren, Hannes</creator><creator>Forssen, Per-Erik</creator><general>IEEE</general><scope>6IE</scope><scope>6IH</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIO</scope><scope>ADTPV</scope><scope>BNKNJ</scope><scope>DG8</scope></search><sort><creationdate>201806</creationdate><title>Spline Error Weighting for Robust Visual-Inertial Fusion</title><author>Ovren, Hannes ; Forssen, Per-Erik</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-i213t-c1f3f54fa8c4181cbdfa073971f8e493014efa7c4f3e7be5ff3b2e4bdcbc9f863</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>2018</creationdate><topic>Approximation error</topic><topic>Cameras</topic><topic>Discrete Fourier transforms</topic><topic>Splines (mathematics)</topic><topic>Trajectory</topic><topic>Videos</topic><toplevel>online_resources</toplevel><creatorcontrib>Ovren, Hannes</creatorcontrib><creatorcontrib>Forssen, Per-Erik</creatorcontrib><collection>IEEE Electronic Library (IEL) Conference Proceedings</collection><collection>IEEE Proceedings Order Plan (POP) 1998-present by volume</collection><collection>IEEE Xplore All Conference Proceedings</collection><collection>IEEE Xplore</collection><collection>IEEE Proceedings Order Plans (POP) 1998-present</collection><collection>SwePub</collection><collection>SwePub Conference</collection><collection>SWEPUB Linköpings universitet</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Ovren, Hannes</au><au>Forssen, Per-Erik</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>Spline Error Weighting for Robust Visual-Inertial Fusion</atitle><btitle>2018 IEEE CVF Conference on Computer Vision and Pattern Recognition</btitle><stitle>CVPR</stitle><date>2018-06</date><risdate>2018</risdate><spage>321</spage><epage>329</epage><pages>321-329</pages><eissn>1063-6919</eissn><isbn>9781538664216</isbn><isbn>1538664216</isbn><isbn>9781538664209</isbn><isbn>1538664208</isbn><eisbn>9781538664209</eisbn><eisbn>1538664208</eisbn><coden>IEEPAD</coden><abstract>In this paper we derive and test a probability-based weighting that can balance residuals of different types in spline fitting. In contrast to previous formulations, the proposed spline error weighting scheme also incorporates a prediction of the approximation error of the spline fit. We demonstrate the effectiveness of the prediction in a synthetic experiment, and apply it to visual-inertial fusion on rolling shutter cameras. This results in a method that can estimate 3D structure with metric scale on generic first-person videos. We also propose a quality measure for spline fitting, that can be used to automatically select the knot spacing. Experiments verify that the obtained trajectory quality corresponds well with the requested quality. Finally, by linearly scaling the weights, we show that the proposed spline error weighting minimizes the estimation errors on real sequences, in terms of scale and end-point errors.</abstract><pub>IEEE</pub><doi>10.1109/CVPR.2018.00041</doi><tpages>9</tpages></addata></record>
fulltext fulltext_linktorsrc
identifier ISBN: 9781538664216
ispartof 2018 IEEE CVF Conference on Computer Vision and Pattern Recognition, 2018, p.321-329
issn 1063-6919
language eng
recordid cdi_ieee_primary_8578139
source IEEE Xplore All Conference Series
subjects Approximation error
Cameras
Discrete Fourier transforms
Splines (mathematics)
Trajectory
Videos
title Spline Error Weighting for Robust Visual-Inertial Fusion
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-10T03%3A05%3A09IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-swepub_CHZPO&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=proceeding&rft.atitle=Spline%20Error%20Weighting%20for%20Robust%20Visual-Inertial%20Fusion&rft.btitle=2018%20IEEE%20CVF%20Conference%20on%20Computer%20Vision%20and%20Pattern%20Recognition&rft.au=Ovren,%20Hannes&rft.date=2018-06&rft.spage=321&rft.epage=329&rft.pages=321-329&rft.eissn=1063-6919&rft.isbn=9781538664216&rft.isbn_list=1538664216&rft.isbn_list=9781538664209&rft.isbn_list=1538664208&rft.coden=IEEPAD&rft_id=info:doi/10.1109/CVPR.2018.00041&rft.eisbn=9781538664209&rft.eisbn_list=1538664208&rft_dat=%3Cswepub_CHZPO%3Eoai_DiVA_org_liu_149495%3C/swepub_CHZPO%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-i213t-c1f3f54fa8c4181cbdfa073971f8e493014efa7c4f3e7be5ff3b2e4bdcbc9f863%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_id=info:pmid/&rft_ieee_id=8578139&rfr_iscdi=true