Estimation of Absolute Scale in Monocular SLAM Using Synthetic Data
This paper addresses the problem of scale estimation in monocular SLAM by estimating absolute distances between camera centers of consecutive image frames. These estimates would improve the overall performance of classical (not deep) SLAM systems and allow metric feature locations to be recovered fr...
Saved in:
Published in | 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW) pp. 803 - 812 |
---|---|
Main Authors | , , , , |
Format | Conference Proceeding |
Language | English |
Published |
IEEE
01.10.2019
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Abstract | This paper addresses the problem of scale estimation in monocular SLAM by estimating absolute distances between camera centers of consecutive image frames. These estimates would improve the overall performance of classical (not deep) SLAM systems and allow metric feature locations to be recovered from a single monocular camera. We propose several network architectures that lead to an improvement of scale estimation accuracy over the state of the art. In addition, we exploit a possibility to train the neural network only with synthetic data derived from a computer graphics simulator. Our key insight is that, using only synthetic training inputs, we can achieve similar scale estimation accuracy as that obtained from real data. This fact indicates that fully annotated simulated data is a viable alternative to existing deep-learning-based SLAM systems trained on real (unlabeled) data. Our experiments with unsupervised domain adaptation also show that the difference in visual appearance between simulated and real data does not affect scale estimation results. Our method operates with low-resolution images (0.03 MP), which makes it practical for real-time SLAM applications with a monocular camera. |
---|---|
AbstractList | This paper addresses the problem of scale estimation in monocular SLAM by estimating absolute distances between camera centers of consecutive image frames. These estimates would improve the overall performance of classical (not deep) SLAM systems and allow metric feature locations to be recovered from a single monocular camera. We propose several network architectures that lead to an improvement of scale estimation accuracy over the state of the art. In addition, we exploit a possibility to train the neural network only with synthetic data derived from a computer graphics simulator. Our key insight is that, using only synthetic training inputs, we can achieve similar scale estimation accuracy as that obtained from real data. This fact indicates that fully annotated simulated data is a viable alternative to existing deep-learning-based SLAM systems trained on real (unlabeled) data. Our experiments with unsupervised domain adaptation also show that the difference in visual appearance between simulated and real data does not affect scale estimation results. Our method operates with low-resolution images (0.03 MP), which makes it practical for real-time SLAM applications with a monocular camera. |
Author | Mouritzen, Daniel Kaestner, Ralf Rukhovich, Danila Rufli, Martin Velizhev, Alexander |
Author_xml | – sequence: 1 givenname: Danila surname: Rukhovich fullname: Rukhovich, Danila organization: IBM Research - Zurich – sequence: 2 givenname: Daniel surname: Mouritzen fullname: Mouritzen, Daniel organization: IBM Research - Zurich – sequence: 3 givenname: Ralf surname: Kaestner fullname: Kaestner, Ralf organization: IBM Research - Zurich – sequence: 4 givenname: Martin surname: Rufli fullname: Rufli, Martin organization: IBM Research - Zurich – sequence: 5 givenname: Alexander surname: Velizhev fullname: Velizhev, Alexander organization: IBM Research - Zurich |
BookMark | eNotjrFOwzAURQ0CiVKyI7H4B1Ke37OdeIxCgUqtGEphrJzEAaNgo8Yd-vdEguWe5ejoXrOLEINj7FbAQggw96u6fntfIAizABBQnrHMFKUosBQKkMw5m6EsKDdGyiuWjeMXTJ4WyhDMWL0ck_-2ycfAY8-rZozDMTm-be3guA98E0Nsj4M98O262vDd6MMH355C-nTJt_zBJnvDLns7jC7755ztHpev9XO-fnla1dU69wiU8qbUaBrqZNtQ0WtX2LYvCGkaJRVq0gYbAFM6SeRaiU1HwpZWS2U0dj3N2d1f1zvn9j-H6ffhtDeAqJSkX5XdS0w |
CODEN | IEEPAD |
ContentType | Conference Proceeding |
DBID | 6IE 6IL CBEJK RIE RIL |
DOI | 10.1109/ICCVW.2019.00108 |
DatabaseName | IEEE Electronic Library (IEL) Conference Proceedings IEEE Proceedings Order Plan All Online (POP All Online) 1998-present by volume IEEE Xplore All Conference Proceedings IEEE Xplore IEEE Proceedings Order Plans (POP All) 1998-Present |
DatabaseTitleList | |
Database_xml | – sequence: 1 dbid: RIE name: IEEE Xplore url: https://proxy.k.utb.cz/login?url=https://ieeexplore.ieee.org/ sourceTypes: Publisher |
DeliveryMethod | fulltext_linktorsrc |
Discipline | Applied Sciences |
EISBN | 9781728150239 172815023X |
EISSN | 2473-9944 |
EndPage | 812 |
ExternalDocumentID | 9022554 |
Genre | orig-research |
GroupedDBID | 6IE 6IF 6IH 6IK 6IL 6IM 6IN AAJGR ABLEC ADZIZ ALMA_UNASSIGNED_HOLDINGS BEFXN BFFAM BGNUA BKEBE BPEOZ CBEJK CHZPO IEGSK IPLJI OCL RIE RIL RNS |
ID | FETCH-LOGICAL-i203t-b8629b3d4cb37f6e7acf7323f73545263692b0098e433ec42bd31a8a645962df3 |
IEDL.DBID | RIE |
IngestDate | Wed Jun 26 19:27:09 EDT 2024 |
IsPeerReviewed | false |
IsScholarly | false |
Language | English |
LinkModel | DirectLink |
MergedId | FETCHMERGED-LOGICAL-i203t-b8629b3d4cb37f6e7acf7323f73545263692b0098e433ec42bd31a8a645962df3 |
PageCount | 10 |
ParticipantIDs | ieee_primary_9022554 |
PublicationCentury | 2000 |
PublicationDate | 2019-Oct |
PublicationDateYYYYMMDD | 2019-10-01 |
PublicationDate_xml | – month: 10 year: 2019 text: 2019-Oct |
PublicationDecade | 2010 |
PublicationTitle | 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW) |
PublicationTitleAbbrev | ICCVW |
PublicationYear | 2019 |
Publisher | IEEE |
Publisher_xml | – name: IEEE |
SSID | ssj0001615930 |
Score | 1.7919226 |
Snippet | This paper addresses the problem of scale estimation in monocular SLAM by estimating absolute distances between camera centers of consecutive image frames.... |
SourceID | ieee |
SourceType | Publisher |
StartPage | 803 |
SubjectTerms | Absolute Cameras Estimation Scale Simultaneous localization and mapping SLAM structure from motion Three-dimensional displays Training Visual odometry Visualization |
Title | Estimation of Absolute Scale in Monocular SLAM Using Synthetic Data |
URI | https://ieeexplore.ieee.org/document/9022554 |
hasFullText | 1 |
inHoldings | 1 |
isFullTextHit | |
isPrint | |
link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwjV1NS8NAEB3anjxVbcVv9uDRtElmm49jiS1VrAi12lvJJhMoQiKaHvTXO7tJWxEPXsISAllmdnizu2_mAVx5UjlO4ihLNyqxZDwIOOZ8tGSi0sxGxuxQ1w5PH7zJXN4tBosGXG9rYYjIkM-op4fmLj8tkrU-KuuHDDgMf01oBrZb1WrtzlMYmkO0NzeRdti_jaLnF03eMh0ptX7kD_0UAx_jNkw3P65YI6-9dal6ydevnoz_ndk-dHeFeuJxC0EH0KD8ENp1ZinquP3oQDTiSK6KFEWRiaEyK474C8YHscoFh3ZhGKlidj-cCkMkELPPnNNDXlniJi7jLszHo6doYtXyCdbKtbG0FG9WQoUpmx39zCM_TjIfXeSHERZHL3SV7idKEpES6aoUnTiItdc8N83wCFp5kdMxCL2JwiCVpDhbxMTTCuWcWNiEFCpSwQl0tE2Wb1WHjGVtjtO_X5_BnvZKRYk7h1b5vqYLhvZSXRqffgPHj6Fu |
link.rule.ids | 310,311,786,790,795,796,802,27956,55107 |
linkProvider | IEEE |
linkToHtml | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwjV1NT8JAEJ0gHvSECsZv9-DRQtspbfdIKgaUEhNAuZFuu02ISWu0HPTXO7stYIwHL82madLNzE7e7O6beQA3riMsK7aEoRqVGE7U9SnmPDScWCSpiYTZXNUOh2N3MHMe5t15DW43tTBSSk0-k2011Hf5SR6v1FFZhxPgEPztwC7hvOmV1VrbExUCZ47m-i7S5J1hEDy_KPqW7kmpFCR_KKhoALlvQLj-dckbeW2vCtGOv351Zfzv3A6gtS3VY08bEDqEmsyOoFHllqyK3I8mBH2K5bJMkeUp6wm95iR9QQjBlhmj4M41J5VNRr2QaSoBm3xmlCDS2mJ3URG1YHbfnwYDoxJQMJa2iYUhaLvCBSZkePRSV3pRnHpoIz20tDi63Baqo6h0EGXs2CJBK_Ij5TfXTlI8hnqWZ_IEmNpGoZ84UlC-iLGrNMoptTAlSi6k8E-hqWyyeCt7ZCwqc5z9_foa9gbTcLQYDceP57CvPFQS5C6gXryv5CUBfSGutH-_AXwLpMI |
openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=proceeding&rft.title=2019+IEEE%2FCVF+International+Conference+on+Computer+Vision+Workshop+%28ICCVW%29&rft.atitle=Estimation+of+Absolute+Scale+in+Monocular+SLAM+Using+Synthetic+Data&rft.au=Rukhovich%2C+Danila&rft.au=Mouritzen%2C+Daniel&rft.au=Kaestner%2C+Ralf&rft.au=Rufli%2C+Martin&rft.date=2019-10-01&rft.pub=IEEE&rft.eissn=2473-9944&rft.spage=803&rft.epage=812&rft_id=info:doi/10.1109%2FICCVW.2019.00108&rft.externalDocID=9022554 |